{"id":40397,"date":"2023-02-03T22:16:18","date_gmt":"2023-02-03T14:16:18","guid":{"rendered":"https:\/\/swarma.org\/?p=40397"},"modified":"2023-02-03T22:16:18","modified_gmt":"2023-02-03T14:16:18","slug":"sci-adv-%e7%bb%bc%e8%bf%b0%ef%bc%9a%e6%9c%ba%e5%99%a8%e5%ad%a6%e4%b9%a0%e6%96%b9%e6%b3%95%e5%9c%a8%e7%a4%be%e4%bc%9a%e5%92%8c%e5%81%a5%e5%ba%b7%e7%a7%91%e5%ad%a6%e4%b8%ad%e7%9a%84%e5%ba%94%e7%94%a8","status":"publish","type":"post","link":"https:\/\/swarma.org\/?p=40397","title":{"rendered":"Sci. Adv. \u7efc\u8ff0\uff1a\u673a\u5668\u5b66\u4e60\u65b9\u6cd5\u5728\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u4e2d\u7684\u5e94\u7528"},"content":{"rendered":"
\n
\n
\n
\n

<\/p>\n<\/section>\n

\n
\n
\n
\n
\n
\n

<\/section>\n<\/section>\n
\n
\n

\u6458\u8981<\/strong><\/p>\n<\/section>\n<\/section>\n

\n

<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n
\n
\n
\n

\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u4e2d\u4f7f\u7528\u7684\u673a\u5668\u5b66\u4e60\uff08Machine Learning, ML\uff09\u65b9\u6cd5\u9700\u6ee1\u8db3\u63cf\u8ff0\u3001\u9884\u6d4b\u3001\u56e0\u679c\u63a8\u65ad\u7b49\u9884\u671f\u7814\u7a76\u76ee\u7684\u3002\u672c\u6587\u901a\u8fc7\u5c06\u7edf\u8ba1\u5206\u6790\u5f15\u5165\u8fd9\u4e9b\u5b66\u79d1\uff0c\u5bf9\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u4e2d\u7684\u7814\u7a76\u95ee\u9898\u4e0eML\u65b9\u6cd5\u8fdb\u884c\u4e86\u5168\u9762\u3001\u7cfb\u7edf\u7684\u603b\u7ed3\u3002\u6211\u4eec\u5c06\u5176\u5206\u7c7b\u4e3a\u63cf\u8ff0\u3001\u9884\u6d4b\u3001\u53cd\u4e8b\u5b9e\u9884\u6d4b\u548c\u56e0\u679c\u7ed3\u6784\u5b66\u4e60\uff0c\u4f8b\u5982\uff0c\u4f30\u7b97\u4e0d\u826f\u793e\u4f1a\u6216\u5065\u5eb7\u7ed3\u679c\u7684\u53d1\u751f\u7387\u3001\u9884\u6d4b\u4e8b\u4ef6\u7684\u98ce\u9669\u3001\u8bc6\u522b\u98ce\u9669\u56e0\u7d20\u6216\u4e0d\u826f\u7ed3\u679c\u7684\u539f\u56e0\u3001\u89e3\u91ca\u5e38\u89c1\u7684ML\u6027\u80fd\u6307\u6807\u7b49\u3002\u8fd9\u4e00\u5f52\u7eb3\u6709\u52a9\u4e8e\u5728\u8003\u8651\u4e0e\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u76f8\u5173\u9886\u57df\u95ee\u9898\u65f6\uff0c\u5145\u5206\u5229\u7528ML\u7684\u4f18\u52bf\u5e76\u52a0\u5febML\u5e94\u7528\u7684\u666e\u53ca\uff0c\u4ee5\u63a8\u8fdb\u57fa\u7840\u548c\u5e94\u7528\u7684\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u7814\u7a76\u3002<\/span><\/strong><\/span><\/strong><\/p>\n<\/section>\n

\n


<\/p>\n<\/section>\n

\n

\u5173\u952e\u8bcd\uff1a\u673a\u5668\u5b66\u4e60<\/strong><\/strong><\/span><\/strong><\/span><\/strong><\/span><\/strong><\/strong><\/span><\/strong><\/span><\/strong><\/span><\/strong><\/strong><\/strong><\/span><\/strong><\/span><\/strong><\/strong><\/strong><\/strong><\/strong><\/strong><\/strong><\/strong><\/strong><\/strong><\/strong><\/strong><\/strong><\/strong><\/strong><\/strong><\/p>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n

\n
\n
\n
<\/svg><\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n
\n
<\/strong><\/section>\n<\/section>\n
\n

Anja K. Leist, et al.<\/span> <\/span>| \u4f5c\u8005<\/strong><\/span><\/span><\/p>\n

\u8d75\u96e8\u777f<\/span> <\/span>| \u8bd1\u8005<\/span><\/strong><\/span><\/span><\/p>\n

\u90ed\u745e\u4e1c<\/span> <\/span>| \u5ba1\u6821<\/span><\/span><\/strong><\/span><\/span><\/p>\n

\u9093\u4e00\u96ea<\/span> <\/span>| <\/strong><\/span>\u7f16\u8f91<\/strong><\/p>\n


<\/p>\n


<\/p>\n<\/section>\n

\n

<\/p>\n<\/section>\n

\n
\n
\n
\n
\u8bba\u6587\u9898\u76ee\uff1aMapping of machine learning approaches for description, prediction, and causal inference in the social and health sciences<\/span><\/section>\n
\u8bba\u6587\u94fe\u63a5\uff1ahttps:\/\/www.science.org\/doi\/10.1126\/sciadv.abk1942<\/span><\/span><\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/blockquote>\n

<\/section>\n


<\/p>\n

\n
\n
\n
\n
\n
\n
\n
\n
\n
\n
\n
\n
\n
\n
\n
\n
\n
\n
\n
\n
\u76ee\u5f55<\/strong><\/section>\n

<\/section>\n
\n

1. \u5f15\u8a00<\/span><\/strong><\/p>\n

2. \u5c06\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u95ee\u9898\u8f6c\u5316\u4e3a\u673a\u5668\u5b66\u4e60\u670d\u52a1<\/span><\/strong><\/p>\n

3. \u673a\u5668\u5b66\u4e60\u7684\u57fa\u7840<\/span><\/strong><\/p>\n

4. \u63cf\u8ff0\u4e2d\u7684\u673a\u5668\u5b66\u4e60<\/span><\/strong><\/p>\n

5. \u9884\u6d4b\u4e2d\u7684\u673a\u5668\u5b66\u4e60<\/span><\/strong><\/p>\n

6. \u56e0\u679c\u63a8\u65ad\u4e2d\u7684\u673a\u5668\u5b66\u4e60<\/span><\/strong><\/p>\n

7. \u673a\u5668\u5b66\u4e60\u8868\u73b0\u6027\u80fd\u8bc4\u4f30<\/span><\/strong><\/p>\n

8. \u672a\u6765\u5c55\u671b<\/span><\/strong><\/p>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n


<\/p>\n


<\/p>\n

\n
\n
\n
\n
\n

<\/section>\n

<\/section>\n
\n
<\/svg><\/section>\n<\/section>\n<\/section>\n
\n

1. \u5f15\u8a00<\/strong><\/strong><\/p>\n<\/section>\n

\n

<\/section>\n

<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/h3>\n

<\/section>\n
\u4e0d\u540c\u4e8e\u8bf8\u591a\u4f20\u7edf\u7684\u7edf\u8ba1\u65b9\u6cd5\uff0c\u968f\u7740\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u9886\u57df\u4e2d\u5927\u578b\u6570\u636e\u96c6\u7684\u53ef\u7528\u6027\u589e\u52a0\uff0c\u673a\u5668\u5b66\u4e60<\/span>\uff08Machine Learning, ML\uff09<\/span>\u65b9\u6cd5\u6709\u6f5c\u529b\u6539\u5584\u7ecf\u9a8c\u5206\u6790\u7684\u5404\u4e2a\u65b9\u9762\u3002\u8fd9\u5305\u62ec\u4e86\u9884\u6d4b\u9886\u57df\u4e2d\u5927\u91cf\u6570\u636e\u7684\u5feb\u901f\u5904\u7406\u3001\u56e0\u679c\u95f4\u7684\u975e\u7ebf\u6027\u548c\u9ad8\u9636\u5173\u7cfb\u4ee5\u53ca\u9884\u6d4b\u51c6\u786e\u6027\u7684\u63d0\u9ad8\u3002\u4f46\u662f\uff0c\u4ece\u793e\u4f1a\u5b66\u3001\u5fc3\u7406\u5b66\u3001\u7ecf\u6d4e\u5b66\u5230\u793e\u4f1a\u3001\u4e34\u5e8a\u6d41\u884c\u75c5\u5b66\u548c\u516c\u5171\u536b\u751f\u5b66\uff0c\u673a\u5668\u5b66\u4e60\u65b9\u6cd5\u5728\u793e\u4f1a\u548c\u5065\u5eb7\u9886\u57df\u4e2d\u5e94\u7528\u7684\u8fdb\u5c55\u76f8\u5f53\u7f13\u6162\uff0c\u81f3\u4eca\u4ecd\u7136\u662f\u4e0d\u5b8c\u5584\u7684\u3002\u8fd9\u5728\u4e00\u5b9a\u7a0b\u5ea6\u4e0a\u53ef\u5f52\u7eb3\u4e3a\u4e09\u70b9\uff0c\u5373\u7f3a\u4e4f\u5b66\u79d1\u4e4b\u95f4\u6c9f\u901a\u3001\u96be\u4ee5\u5c06\u9886\u57df\u77e5\u8bc6\u7eb3\u5165\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u9886\u57df\u4e2d\u7684\u5206\u6790\u65b9\u6cd5\u4ee5\u53ca\u7f3a\u4e4f\u9002\u5408\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u9886\u57df\u7684\u673a\u5668\u5b66\u4e60\u65b9\u6cd5\u6982\u8ff0\u3002<\/span><\/section>\n

<\/span><\/section>\n
\u672c\u6587\u65e8\u5728\u5c06\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u7684\u7814\u7a76\u76ee\u6807\u7cfb\u7edf\u5730\u6620\u5c04\u5230\u9002\u5f53\u7684\u673a\u5668\u5b66\u4e60\u65b9\u6cd5\uff0c\u8fdb\u800c\u63d0\u4f9b\u4e00\u4e2a\u9ad8\u6c34\u5e73\u7684\u3001\u975e\u6280\u672f\u6027\u7684\u673a\u5668\u5b66\u4e60\u65b9\u6cd5\u5de5\u5177\u7bb1\u3001\u89e3\u91ca\u673a\u5668\u5b66\u4e60\u4e2d\u7684\u5e38\u89c1\u5ea6\u91cf\u6807\u51c6\u5e76\u6307\u5bfc\u7814\u7a76\u4eba\u5458\u89e3\u51b3\u673a\u5668\u5b66\u4e60\u5efa\u6a21\u4e2d\u7684\u5e38\u89c1\u95ee\u9898\u3002\u672c\u6587\u4fa7\u91cd\u4e8e\u7814\u7a76\u95ee\u9898\uff0c\u6d89\u53ca\u4e86\u4ee5\u4eba\u7c7b\u4e3a\u7814\u7a76\u5355\u4f4d\u7684\u6570\u636e\u96c6\u4ee5\u53ca\u4e34\u5e8a\u8bc4\u4f30\u6216\u81ea\u6211\u62a5\u544a\u3002\u9488\u5bf9\u5927\u591a\u6570\u4f7f\u7528\u673a\u5668\u5b66\u4e60\u7684\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u7684\u7814\u7a76\u4e2d\uff0c\u672c\u6587\u91c7\u7528\u5728\u9759\u6001\u6570\u636e\u96c6\u4e0a\u8bad\u7ec3\u7684\u6a21\u578b\uff0c\u5373\u6a21\u578b\u4e0d\u662f\u8fde\u7eed\u5904\u7406\u65b0\u7684\u6570\u636e\uff0c\u800c\u662f\u4f9d\u8d56\u4e8e\u961f\u5217\u7814\u7a76\u6216\u6570\u636e\u6536\u96c6\u548c\u6e05\u7406\u7ed3\u675f\u540e\u7684\u8c03\u67e5\u7684\u6709\u9650\u6570\u636e\u96c6\u3002<\/span><\/section>\n

<\/span><\/section>\n
\u672c\u6587\u7efc\u8ff0\u5e94\u8be5\u88ab\u770b\u4f5c\u662f\u5bf9\u6d41\u884c\u75c5\u5b66\u548c\u5065\u5eb7\u7814\u7a76(1) \uff0c\u5fc3\u7406\u5b66(2)\u548c\u7ecf\u6d4e\u5b66(3)\u7b49\u9886\u57df\u7684\u673a\u5668\u5b66\u4e60\u5bfc\u8bba\u7684\u8865\u5145\u3002\u5bf9\u4e8e\u4e00\u822c\u7684\u7edf\u8ba1\u5b66\u4e60\u4ecb\u7ecd\uff0c\u611f\u5174\u8da3\u7684\u8bfb\u8005\u53ef\u4ee5\u53c2\u8003\u5173\u4e8e\u8fd9\u4e9b\u65b9\u6cd5\u7684\u4f18\u79c0\u6559\u79d1\u4e66(4,5)\u3002<\/span><\/section>\n

<\/span><\/section>\n
\u672c\u6587\u7684\u5176\u4f59\u5185\u5bb9\u7ec4\u7ec7\u5982\u4e0b\uff1a\u7b2c\u4e8c\u8282\u90e8\u5206\u6982\u8ff0\u4e86\u5c06\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u7684\u7814\u7a76\u76ee\u7684\u6620\u5c04\u5230\u9002\u5f53\u7684\u673a\u5668\u5b66\u4e60\u65b9\u6cd5\u7684\u4efb\u52a1\u3002\u7b2c\u4e09\u8282\u6db5\u76d6\u4e86\u90e8\u5206\u673a\u5668\u5b66\u4e60\u7684\u57fa\u7840\u77e5\u8bc6\uff0c\u7279\u522b\u662f\u4f20\u7edf\u7684\u673a\u5668\u5b66\u4e60\u5206\u7c7b\u3001\u6570\u636e\u51c6\u5907\u3001\u6a21\u578b\u5efa\u7acb\u4ee5\u53ca\u673a\u5668\u5b66\u4e60\u7684\u201c\u73b0\u5b9e\u4e16\u754c\u201d\u5e94\u7528\u3002\u7b2c\u56db\u8282\u5230\u7b2c\u516d\u8282\u63cf\u8ff0\u4e86\u673a\u5668\u5b66\u4e60\u65b9\u6cd5\u5bf9\u63cf\u8ff0\u3001\u9884\u6d4b\u548c\u56e0\u679c\u63a8\u7406\u7814\u7a76\u76ee\u7684\u7684\u6620\u5c04\uff0c\u5e76\u7ed9\u51fa\u4e86\u7ecf\u9a8c\u5b9e\u4f8b\u3002\u7b2c\u4e03\u8282\u7ed9\u51fa\u4e86\u673a\u5668\u5b66\u4e60\u6027\u80fd\u6307\u6807\u7684\u6982\u8ff0\u3002\u7b2c\u516b\u8282\u4ee5\u5c55\u671b\u7ed3\u675f\u5168\u6587\u3002<\/span><\/section>\n

<\/section>\n

\n


<\/p>\n<\/h2>\n

\n
\n
\n
\n
\n

<\/section>\n

<\/section>\n
\n
<\/svg><\/section>\n<\/section>\n<\/section>\n
\n

2. \u5c06\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u95ee\u9898<\/strong><\/strong><\/p>\n

\u8f6c\u5316\u4e3a\u673a\u5668\u5b66\u4e60\u4efb\u52a1<\/strong><\/strong><\/p>\n<\/section>\n

\n

<\/section>\n

<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/h3>\n

<\/section>\n
\u7b80\u800c\u8a00\u4e4b\uff0c\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u4e2d\u7684\u5e38\u89c1\u7814\u7a76\u76ee\u7684\u53ef\u4ee5\u6839\u636e\u7814\u7a76\u4eba\u5458\u7684\u610f\u56fe\u5f52\u7c7b\u4e3a: (i)\u63cf\u8ff0\u73b0\u8c61\uff0c(ii)\u9884\u6d4b\u793e\u4f1a\u6216\u5065\u5eb7\u7ed3\u679c\uff0c\u4ee5\u53ca(iii)\u627e\u5230\u539f\u56e0\u4ee5\u53ca\u53ef\u80fd\u5e72\u9884\u7ed3\u679c\u7684\u539f\u56e0\u3002\u867d\u7136\u4e0d\u662f\u6240\u6709\u7684\u7814\u7a76\u95ee\u9898\u90fd\u53ef\u4ee5\u6309\u7167\u8fd9\u4e00\u5f52\u7eb3\u8fdb\u884c\u4e25\u683c\u7684\u533a\u5206\uff0c\u672c\u6587\u4ecd\u8be6\u7ec6\u5730\u4ecb\u7ecd\u4e0e\u63cf\u8ff0\u3001\u9884\u6d4b\u548c\u56e0\u679c\u63a8\u7406\u76f8\u5173\u7684\u5177\u4f53\u7814\u7a76\u95ee\u9898(6)\u3002\u672c\u6587\u5c06\u8fd9\u4e9b\u7814\u7a76\u95ee\u9898\u6620\u5c04\u5230\u9002\u5f53\u7684\u673a\u5668\u5b66\u4e60\u65b9\u6cd5\u4e2d\uff0c\u5e76\u5728\u53ef\u80fd\u7684\u60c5\u51b5\u4e0b\u4f7f\u7528\u5b9e\u8bc1\u7814\u7a76\u4f5c\u4e3a\u4f8b\u8bc1\u3002<\/span><\/section>\n

<\/span><\/section>\n
\u672c\u6587\u6240\u603b\u7ed3\u7684ML \u65b9\u6cd5\u4e0d\u540c\u4e8e\u6570\u636e\u5206\u6790\u4e2d\u7684\u4f20\u7edf\u65b9\u6cd5\uff0c\u4f8b\u5982\uff0c\u63a8\u65ad\u7edf\u8ba1\u5b66\u3001\u7edf\u8ba1\u5b66\u4e60\u4ee5\u53ca\u8ba1\u7b97\u79d1\u5b66\u3002\u5b83\u4eec\u7684\u5171\u540c\u70b9\u662f\u5177\u5907\u5904\u7406\u5927\u91cf\u6570\u636e\u7684\u80fd\u529b\uff0c\u5e76\u4e14\u5efa\u6a21\u548c\u6a21\u578b\u9009\u62e9\u66f4\u591a\u5730\u4f9d\u8d56\u6570\u636e\u7ed3\u6784<\/span>\uff08\u6570\u636e\u9a71\u52a8\uff09<\/span>\uff0c\u800c\u4e0d\u662f\u4f20\u7edf\u7684\u63a8\u65ad\u7edf\u8ba1\u3002<\/span><\/section>\n

<\/span><\/section>\n
\u5f53\u673a\u5668\u5b66\u4e60\u5e94\u7528\u4e8e\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u6240\u7814\u7a76\u7684\u95ee\u9898\u65f6\uff0c\u7edf\u8ba1\u5b66\u77e5\u8bc6\u548c\u9886\u57df\u77e5\u8bc6\u90fd\u662f\u76f8\u5173\u7684\u3002\u63a2\u7d22\u6027\u6570\u636e\u5206\u6790(7)\u88ab\u7528\u4e8e\u603b\u7ed3\u548c\u53ef\u89c6\u5316\u6570\u636e\u4e3b\u8981\u7279\u5f81\uff0c\u662f\u6570\u636e\u5206\u6790\u548c\u673a\u5668\u5b66\u4e60\u6d41\u7a0b\u4e2d\u91cd\u8981\u7684\u7b2c\u4e00\u6b65\u3002\u7136\u800c\uff0c\u4e0d\u53ef\u77e5\u6570\u636e\u63a2\u7d22\uff0c\u5373\u4e0d\u8003\u8651\u9886\u57df\u77e5\u8bc6\u7684\u6570\u636e\u5206\u6790\uff0c\u5728\u5927\u591a\u6570\u60c5\u51b5\u4e0b\u53ea\u80fd\u63d0\u4f9b\u8f83\u5c11\u7684\u89c1\u89e3\u3002\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u9886\u57df\u77e5\u8bc6\u91cd\u8981\u6027\u7684\u4e00\u4e9b\u4f8b\u5b50\u5982\u4e0b:<\/span><\/section>\n

<\/span><\/section>\n
(1)\u51b3\u7b56\u6811\u548c\u5176\u4ed6\u673a\u5668\u5b66\u4e60\u7b97\u6cd5\u4e2d\u8fde\u7eed\u53d8\u91cf\u7684\u504f\u597d\u53ef\u80fd\u5bfc\u81f4\u9ad8\u4f30\u5e74\u9f84\u4f5c\u4e3a\u9884\u6d4b\u56e0\u5b50\uff0c\u7136\u800c\u5b9e\u9645\u4e0a\uff0c\u5e74\u9f84\u4e0e\u793e\u4f1a\u6216\u5065\u5eb7\u73b0\u8c61\u7684\u8054\u7cfb\u8981\u5c11\u5f97\u591a<\/span>\uff08\u4f8b<\/span>\u5982\uff0c\u6574\u4e2a\u751f\u547d\u8fc7\u7a0b\u4e2d\u5728\u8ba4\u77e5\u8868\u73b0\u4e0a\u7684\u5f3a\u70c8\u5f02\u8d28\u6027\uff09<\/span>\u3002<\/span><\/section>\n

<\/span><\/section>\n
(2)\u7f3a\u5931\u7684\u6570\u636e\u53ef\u80fd\u5728\u6570\u636e\u9a71\u52a8\u7684\u5206\u6790\u4e2d\u88ab\u786e\u5b9a\u4e3a\u6709\u610f\u4e49\u7684\u4fe1\u606f<\/span>\uff08\u901a\u5e38\u5982\u6b64\uff09<\/span>\u3002\u7136\u800c\uff0c\u8fd9\u9700\u8981\u5c06\u89c2\u5bdf\u5230\u7684\u4eba\u7fa4\u4ee5\u53ca\u5176\u611f\u5174\u8da3\u7684\u7ed3\u679c\u8fdb\u884c\u80cc\u666f\u5316\u3002<\/span><\/section>\n

<\/span><\/section>\n
(3)\u5728\u6570\u636e\u9a71\u52a8\u5206\u6790\u4e2d\u88ab\u786e\u5b9a\u4e3a\u76f8\u5173\u7684\u7c7b\u522b\u5728\u6982\u5ff5\u4e0a\u53ef\u5177\u6709\u4e0d\u540c\u7684\u610f\u4e49\uff0c\u4f8b\u5982\uff0c\u6559\u80b2\u76f8\u5173\u7684\u7c7b\u522b\u201c\u5176\u4ed6\u201d\u6216\u201c\u90fd\u4e0d\u662f\u201d\u53ef\u80fd\u53cd\u6620\u4e86\u5f02\u5e38\u4f4e\u7684\u6559\u80b2\u6c34\u5e73\u6216\u5728\u56fd\u5916\u83b7\u5f97\u7684\u6559\u80b2\u5b66\u4f4d\uff0c\u4f46\u8fd9\u9700\u8981\u5177\u4f53\u5316\u3002<\/span><\/section>\n

<\/span><\/section>\n
(4)\u5177\u6709\u4e0d\u540c\u793e\u4f1a\u4eba\u53e3\u5b66\u548c\u793e\u4f1a\u7ecf\u6d4e\u5b66\u80cc\u666f\u7684\u4e2a\u4f53\u5728\u884c\u4e3a<\/span>\uff08\u4ee5\u53ca\u8c03\u67e5\u53c2\u4e0e\uff09<\/span>\u4e0a\u7684\u7cfb\u7edf\u5dee\u5f02\u53ef\u80fd\u5bfc\u81f4\u51b3\u5b9a\u56e0\u7d20-\u7ed3\u679c\u5173\u8054\u5b58\u5728\u504f\u5dee\uff0c\u8fd9\u4e0e\u5065\u5eb7\u7684\u793e\u4f1a\u51b3\u5b9a\u56e0\u7d20\u7684\u7814\u7a76\u9ad8\u5ea6\u76f8\u5173\u3002<\/span><\/section>\n

<\/span><\/section>\n
\u6211\u4eec\u6ce8\u610f\u5230\uff0c\u6700\u540e\u4e00\u4e2a\u4f8b\u5b50\u5f80\u5f80\u610f\u5473\u7740\u533b\u7597\u4fdd\u5065\u4e2d\u7684\u4e0d\u5e73\u7b49\u3002\u5c11\u6570\u7fa4\u4f53\u7684\u6570\u636e\u4e0d\u5e73\u8861\u548c\u7ed3\u6784\u4e0d\u5e73\u8861\u53ef\u80fd\u5bfc\u81f4\u4e0d\u5e73\u7b49\u7684 ML \u7b97\u6cd5\uff0c\u8fd9\u6709\u53ef\u80fd\u8fdb\u4e00\u6b65\u52a0\u5267\u5065\u5eb7\u4e0d\u5e73\u7b49(8)\u3002\u867d\u7136\u8fd9\u91cc\u63d0\u51fa\u7684\u4e00\u4e9b\u673a\u5668\u5b66\u4e60\u65b9\u6cd5\uff0c\u7279\u522b\u662f\u7528\u4e8e\u56e0\u679c\u63a8\u7406\u7684\u673a\u5668\u5b66\u4e60\uff0c\u6bd4\u5176\u4ed6\u65b9\u6cd5\u9700\u8981\u66f4\u591a\u7684\u9886\u57df\u77e5\u8bc6\u3002\u4f46\u6211\u4eec\u8ba4\u4e3a\uff0c\u5728\u6240\u6709\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u7684\u7814\u7a76\u95ee\u9898\u4e2d\uff0c\u5927\u91cf\u7684\u9886\u57df\u77e5\u8bc6\u53ef\u4ee5\u6709\u610f\u4e49\u5730\u8d21\u732e\u9886\u57df\uff0c\u8fd9\u662f\u53ef\u89e3\u91ca\u6027\u7684\u5148\u51b3\u6761\u4ef6(9)\u3002\u867d\u7136\u8ba1\u7b97\u9886\u57df\u4f20\u7edf\u4e0a\u5f3a\u8c03\u9884\u6d4b<\/span>\uff08\u4ee5\u53ca\u7edf\u8ba1\u77e5\u8bc6\uff09<\/span>\u7684\u6539\u8fdb\uff0c\u4f46\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u5f80\u5f80\u4f18\u5148\u8003\u8651\u89e3\u91ca<\/span>\uff08\u4ee5\u53ca\u9886\u57df\u77e5\u8bc6\uff09<\/span>(2) \u3002\u672c\u6587\u65e8\u5728\u8868\u660e\uff0c\u63a8\u8fdb\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u7684\u79d1\u5b66\u77e5\u8bc6\u662f\u5fc5\u8981\u7684\u3002<\/span><\/section>\n

<\/span><\/section>\n
\u6709\u5fc5\u8981\u5728\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u7684\u7814\u7a76\u4eba\u5458\u548c\u7ecf\u8fc7\u65b9\u6cd5\u8bad\u7ec3\u7684\u7814\u7a76\u4eba\u5458\u4e4b\u95f4\u5efa\u7acb\u4e00\u4e2a\u6d41\u7545\u7684\u5bf9\u8bdd\uff0c\u4ee5\u907f\u514d\u201c\u91cd\u65b0\u53d1\u660e\u8f66\u8f6e\u201d\u673a\u5668\u5b66\u4e60\u7684\u7814\u7a76\u4eba\u5458\u53ef\u80fd\u7f3a\u4e4f\u9886\u57df\u77e5\u8bc6\uff0c\u5ffd\u89c6\u4e86\u4ee5\u524d\u53d1\u73b0\u7684\u4e0e\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u9ad8\u5ea6\u76f8\u5173\u7684\u6570\u636e\u7279\u5f81<\/span>\uff08\u4f8b\u5982\uff0c\u5728\u6574\u5408\u6570\u636e\u96c6\u65f6\u5bf9\u67d0\u4e9b\u53d8\u91cf\u7684\u91cd\u65b0\u7f16\u7801\u8fc7\u4e8e\u7b80\u5316\uff09<\/span>\u3002\u76f8\u6bd4\u4e4b\u4e0b\uff0c\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u7684\u7814\u7a76\u4eba\u5458\u53ef\u80fd\u6ca1\u6709\u610f\u8bc6\u5230\u7b97\u6cd5\u80cc\u540e\u7684\u590d\u6742\u6570\u5b66\u548c\u7edf\u8ba1\u5b66\u4ee5\u53ca\u6539\u8fdb\u673a\u5668\u5b66\u4e60\u65b9\u6cd5\u7684\u5feb\u901f\u53d1\u5c55\u3002\u603b\u7684\u6765\u8bf4\uff0c\u6839\u636e\u6211\u4eec\u81ea\u5df1\u7684\u7ecf\u9a8c\uff0c\u6211\u4eec\u5efa\u8bae\u8de8\u5b66\u79d1\u7684\u5408\u4f5c\uff0c\u9080\u8bf7\u6570\u636e\u79d1\u5b66\u548c\u673a\u5668\u5b66\u4e60\u4e13\u5bb6\u5728\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u9886\u57df\u505a\u7814\u7a76\uff0c\u5e76\u5e0c\u671b\u8fd9\u91cc\u63d0\u51fa\u7684\u6620\u5c04\u5c06\u4fc3\u8fdb\u4e0d\u540c\u5b66\u79d1\u7684\u76f8\u4e92\u7406\u89e3\u3002<\/span><\/section>\n

\n

<\/section>\n


<\/p>\n<\/h2>\n

\n
\n
\n
\n
\n

<\/section>\n

<\/section>\n
\n
<\/svg><\/section>\n<\/section>\n<\/section>\n
\n

3. \u673a\u5668\u5b66\u4e60\u7684\u57fa\u7840<\/strong><\/strong><\/p>\n<\/section>\n

\n

<\/section>\n

<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/h3>\n

<\/section>\n
\u5728\u8f6c\u5411\u673a\u5668\u5b66\u4e60\u8fdb\u884c\u63cf\u8ff0\u3001\u9884\u6d4b\u548c\u56e0\u679c\u63a8\u65ad\u4e4b\u524d\uff0c\u6211\u4eec\u9996\u5148\u7b80\u8981\u4ecb\u7ecd\u4e86\u673a\u5668\u5b66\u4e60\uff0c\u4ece(i)\u4f20\u7edf\u7684\u673a\u5668\u5b66\u4e60\u5206\u7c7b(ii)\u6570\u636e\u51c6\u5907(iii)\u6a21\u578b\u5efa\u7acb\u4ee5\u53ca(iv)\u73b0\u5b9e\u4e16\u754c\u4e2d\u673a\u5668\u5b66\u4e60\u7684\u53ef\u4fe1\u5ea6\u3002<\/span><\/section>\n

<\/span><\/section>\n

\n
3.1 \u4f20\u7edf\u673a\u5668\u5b66\u4e60\u65b9\u6cd5\u7684\u5206\u7c7b<\/span><\/strong><\/span><\/section>\n

<\/span><\/section>\n<\/h3>\n
\u4ece\u6570\u636e\u79d1\u5b66\u7684\u89d2\u5ea6\u6765\u770b\uff0c\u5927\u591a\u6570\u673a\u5668\u5b66\u4e60\u65b9\u6cd5\u53ef\u4ee5\u5206\u4e3a\u65e0\u76d1\u7763\u5f0f\u5b66\u4e60\u3001\u6709\u76d1\u7763\u5f0f\u5b66\u4e60\u548c\u5f3a\u5316\u5b66\u4e60\u4e09\u4e2a\u4e3b\u8981\u5206\u652f:<\/span><\/section>\n

<\/span><\/section>\n
\uff081\uff09\u65e0\u76d1\u7763\u5f0f\u5b66\u4e60\u662f\u6307\u4ece\u672a\u6807\u8bb0\u7684\u6570\u636e\u4e2d\u5b66\u4e60\u6a21\u5f0f\u7684\u7b97\u6cd5\uff0c\u5373\uff0c\u4ece\u672a\u88ab\u4eba\u7c7b\u6807\u8bb0\u7684\u53d8\u91cf\u4e2d\u5b66\u4e60\u6a21\u5f0f\u7684\u7b97\u6cd5\u3002\u4f8b\u5982\uff0c\u65e0\u76d1\u7763\u5f0f\u5b66\u4e60\u5c06\u6839\u636e\u76f8\u4f3c\u6027\u5bf9\u6570\u636e\u6837\u672c\u8fdb\u884c\u5206\u7c7b\u3002<\/span><\/section>\n

<\/span><\/section>\n
\uff082\uff09\u6709\u76d1\u7763\u5f0f\u5b66\u4e60\u65e8\u5728\u901a\u8fc7\u6807\u8bb0\u6570\u636e\u5b66\u4e60\u4e00\u4e2a\u5c06\u8f93\u5165\u6620\u5c04\u5230\u8f93\u51fa\u7684\u51fd\u6570\uff0c\u5373\u4e3a\u8f93\u51fa\u53d8\u91cf\u7684\u7c7b\u522b\u5206\u914d\u6709\u610f\u4e49\u7684\u6807\u8bb0\u6216\u6807\u7b7e\u3002\u5728\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u4e2d\uff0c\u8f93\u5165\u88ab\u79f0\u4e3a\u9884\u6d4b\u56e0\u5b50\u3001\u81ea\u53d8\u91cf\u6216\u66b4\u9732; \u8f93\u51fa\u88ab\u79f0\u4e3a\u7ed3\u679c\u6216\u56e0\u53d8\u91cf\u3002\u6709\u76d1\u7763\u5f0f\u5b66\u4e60\u9700\u8981\u6709\u6807\u7b7e\u7684\u8bad\u7ec3\u6570\u636e\uff0c\u5e76\u4e14\u53ef\u4ee5\u5728\u6709\u6807\u7b7e\u7684\u6d4b\u8bd5\u6570\u636e\u96c6\u4e2d\u8fdb\u884c\u9a8c\u8bc1\u3002\u5c06\u628a\u795e\u7ecf\u7f51\u7edc<\/span>\uff08\u901a\u5e38\u79f0\u4e3a\u4eba\u5de5\u795e\u7ecf\u7f51\u7edc\uff09<\/span>\u4f5c\u4e3a\u9884\u6d4b\u7b97\u6cd5\u7684\u4e00\u4e2a\u4f8b\u5b50\uff0c\u628a\u8d1d\u53f6\u65af\u52a0\u6027\u56de\u5f52\u6811<\/span>\uff08Bayesian Additive Regression Trees, BART\uff09<\/span>\u4f5c\u4e3a\u56e0\u679c\u63a8\u7406\u7684 ML \u7684\u4e00\u4e2a\u4f8b\u5b50\u3002\u60e9\u7f5a\u56de\u5f52\uff0c\u5982\u6700\u5c0f\u7edd\u5bf9\u6536\u7f29\u548c\u9009\u62e9\u7b97\u5b50<\/span>\uff08Least Absolute Shrinkage and Selection Operator, LASSO\uff09<\/span>\u57fa\u4e8e\u5176\u9884\u6d4b\u80fd\u529b\u964d\u4f4e\u4e86\u6a21\u578b\u4e2d\u53d8\u91cf\u7684\u6570\u91cf\uff0c\u5e76\u6709\u52a9\u4e8e\u5f97\u5230\u65e2\u7b80\u7ea6\u53c8\u8868\u73b0\u826f\u597d\u7684\u6a21\u578b\u3002<\/span><\/section>\n

<\/span><\/section>\n
\uff083\uff09\u5f3a\u5316\u5b66\u4e60\u662f\u6307\u667a\u80fd\u4f53\u6839\u636e\u7d2f\u79ef\u62a5\u916c\u7684\u6982\u5ff5\u5bf9\u73af\u5883\u505a\u51fa\u51b3\u7b56\u5e76\u8fdb\u884c\u6539\u8fdb\uff0c\u5373\u667a\u80fd\u4f53\u4f1a\u6839\u636e\u73af\u5883\u7684\u53cd\u9988\u6765\u6539\u53d8\u548c\u4f18\u5316\u8f93\u5165\u3002\u5f3a\u5316\u5b66\u4e60\u53ef\u4ee5\u5e94\u7528\u4e8e\u6570\u636e\u751f\u6210\u7684\u60c5\u51b5\uff0c\u5373\u5728\u5176\u4ed6\u7279\u5f81<\/span>\uff08\u53d8\u91cf\uff09<\/span>\u7684\u63a7\u5236\u4e0b\u64cd\u7eb5\u6cbb\u7597\u53d8\u91cf<\/span>\uff08\u6240\u8c13\u7684 A\/B \u6d4b\u8bd5\uff09<\/span>\u662f\u53ef\u80fd\u7684\u3002<\/span><\/section>\n

<\/span><\/section>\n
\u4f5c\u4e3a\u4e00\u79cd\u5b66\u4e60\u6280\u672f\uff0c\u8fc1\u79fb\u5b66\u4e60\u4f7f\u7528\u591a\u4e2a\u6570\u636e\u6e90\u4ece\u4e00\u4e2a\u6570\u636e\u96c6\u5b66\u4e60\u77e5\u8bc6\uff0c\u5e76\u5c06\u8fd9\u4e9b\u77e5\u8bc6\u8fc1\u79fb\u5230\u5305\u542b\u4e0d\u540c\u4eba\u7fa4\u7684\u53e6\u4e00\u4e2a\u6570\u636e\u96c6\uff0c\u751a\u81f3\u53ef\u80fd\u5305\u542b\u4e0d\u540c\u7684\u7ed3\u679c(10)\u3002\u56e0\u6b64\uff0c\u8fc1\u79fb\u5b66\u4e60\u5c06\u5b66\u5230\u7684\u7279\u5f81\u4ece\u4e00\u79cd\u60c5\u51b5\u8f6c\u79fb\u5230\u53e6\u4e00\u79cd\u60c5\u51b5<\/span>\uff08\u4e00\u81f4\u7684\uff09<\/span>\uff0c\u4ece\u800c\u786e\u5b9a\u5404\u79cd\u60c5\u51b5\u4e0b\u6a21\u5f0f\u548c\u884c\u4e3a\u7684\u5171\u540c\u90e8\u5206\u3002\u56e0\u5176\u7ecf\u5e38\u4e0e\u6709\u6807\u7b7e\u6570\u636e\u4e00\u8d77\u4f7f\u7528\uff0c\u6545\u88ab\u8ba4\u4e3a\u662f\u76d1\u7763\u673a\u5668\u5b66\u4e60\u7684\u4e00\u79cd\u7279\u6b8a\u65b9\u6cd5\u3002\u4f46\u8fc1\u79fb\u5b66\u4e60\u65b9\u6cd5\u5728\u5176\u4ed6\u9886\u57df\u4e5f\u5f97\u5230\u4e86\u4e00\u5b9a\u7684\u53d1\u5c55\uff0c\u4f8b\u5982\uff0c\u5728\u65e0\u76d1\u7763\u5f0f\u5b66\u4e60\u3001\u56fe\u50cf\u8bc6\u522b\u4ee5\u53ca\u672c\u6587\u672a\u63d0\u53ca\u7684\u9886\u57df\u3002\u672c\u6587\u5c06\u63d0\u51fa\u91c7\u7528\u8fc1\u79fb\u5b66\u4e60\u51cf\u5c11\u6570\u636e\u4e0d\u5e73\u8861\u7684\u5f71\u54cd(11)\u3002<\/span><\/section>\n

<\/span><\/section>\n

\n
3.2 \u6570\u636e\u51c6\u5907<\/span><\/strong><\/section>\n

<\/span><\/section>\n<\/h3>\n
\u4e0e\u4f20\u7edf\u7684\u7edf\u8ba1\u5b66\u4e0d\u540c\uff0c\u673a\u5668\u5b66\u4e60\u9700\u8981\u8ba4\u771f\u7684\u6570\u636e\u51c6\u5907\u3002\u5728\u63a5\u4e0b\u6765\u7684\u6587\u7ae0\u5c06\u8ba8\u8bba\u4e0e\u673a\u5668\u5b66\u4e60\u76f8\u5173\u7684\u51e0\u4e2a\u65b9\u9762: (i)\u6570\u636e\u9700\u6c42\u3001(ii)\u7279\u5f81\u9009\u62e9\u4ee5\u53ca(iii)\u7279\u5f81\u5de5\u7a0b\u3002\u4e09\u8005\u4e4b\u4e2d\uff0c\u540e\u4e24\u4e2a\u662f\u5173\u4e8e\u51c6\u5907\u7b97\u6cd5\u7406\u89e3\u7684\u56e0\u679c\u53d8\u91cf\u7684\u65b9\u5f0f(1,12)\u3002<\/span><\/section>\n

<\/span><\/section>\n

\n
3.2.1 \u6570\u636e\u9700\u6c42<\/span><\/strong><\/section>\n

<\/span><\/section>\n<\/h4>\n
\u672c\u6587\u6240\u4ecb\u7ecd\u7684\u5927\u591a\u6570\u673a\u5668\u5b66\u4e60\u65b9\u6cd5\u90fd\u9700\u8981\u6bd4\u4f20\u7edf\u5efa\u6a21\u66f4\u5927\u7684\u6570\u636e\u96c6\uff0c\u4ee5\u671f\u5728\u672a\u8bad\u7ec3\u6570\u636e\u96c6\u4e2d\u8868\u73b0\u6027\u80fd\u4f18\u4e8e\u4f20\u7edf\u5efa\u6a21\u3002\u6570\u636e\u7684\u67d0\u4e9b\u65b9\u9762\u5e94\u8be5\u5177\u6709\u5f88\u5927\u7684\u6570\u91cf\uff0c\u5982\u65f6\u95f4\u70b9\u3001\u53d8\u91cf\u6216\u4e2a\u4f53\u3002\u867d\u7136\u6570\u76ee\u5177\u4f53\u53d6\u51b3\u4e8e\u6570\u636e\u548c\u673a\u5668\u5b66\u4e60\u65b9\u6cd5\uff0c\u4e00\u822c\u800c\u8a00\u662f\u6709\u51e0\u4e07\u4e2a\u6570\u636e\u70b9\u53ef\u7528\u3002\u4f46\u662f\uff0c\u4e00\u4e9b\u5e94\u7528\u7a0b\u5e8f\u53ef\u4ee5\u4f7f\u7528\u975e\u5e38\u5c0f\u7684\u6570\u636e\u96c6\u8fdb\u884c\u63a2\u7d22\u6027\u5206\u6790(1)\u3002\u5982\u540e\u9762\u6240\u8ba8\u8bba\u7684\uff0c\u6240\u9700\u8981\u7684\u6570\u636e\u70b9\u6570\u91cf\u53d6\u51b3\u4e8e\u6570\u636e\u7ef4\u5ea6<\/span>\uff08\u4f8b\u5982\u9884\u6d4b\u56e0\u5b50\u7684\u6570\u91cf\uff09<\/span>\u3001\u6a21\u578b\u590d\u6742\u5ea6\u3001\u672a\u77e5\u771f\u5b9e\u7684\u7a00\u758f\u6027\u4ee5\u53ca\u6570\u636e\u751f\u6210\u673a\u5236\u7684\u590d\u6742\u5ea6(13)\u3002\u56e0\u6b64\uff0cML\u65b9\u6cd5\u53ef\u5e94\u7528\u4e8efat datasets<\/span>\uff08\u7279\u5f81\u7684\u6570\u91cf\u8fdc\u5927\u4e8e\u6837\u672c\u5927\u5c0f\uff09<\/span>\u3002\u8be5\u6570\u636e\u96c6\u4e2d\u6f5c\u5728\u7684\u4fe1\u53f7\u88ab\u8ba4\u4e3a\u662f\u975e\u5e38\u7a00\u758f\u7684\u3002\u4e0e\u975e\u673a\u5668\u5b66\u4e60\u7684\u6570\u636e\u5206\u6790\u7c7b\u4f3c\uff0c\u8ba4\u771f\u5904\u7406\u6570\u636e<\/span>\uff08\u4f8b\u5982\u5728\u534f\u8c03\u8fc7\u7a0b\u4e2d\uff09<\/span>\u5e76\u4e14\u6df1\u5165\u7406\u89e3\u6570\u636e\u6765\u81ea\u54ea\u91cc\u3001\u80fd\u544a\u8bc9\u6211\u4eec\u4ec0\u4e48<\/span>\uff08\u4ee5\u53ca\u4e0d\u80fd\u544a\u8bc9\u6211\u4eec\u4ec0\u4e48\uff09<\/span>\u662f\u81f3\u5173\u91cd\u8981\u7684\u3002\u673a\u5668\u5b66\u4e60\u7684\u5de5\u4f5c\u6d41\u7a0b\u5df2\u7ecf\u5728\u51e0\u7bc7\u4ecb\u7ecd\u6027\u7684\u8bba\u6587\u4e2d\u63cf\u8ff0\u8fc7\uff0c\u8be6\u89c1(1)\u3002\u6ce8\u610f\u5230\u8054\u90a6\u5b66\u4e60\u8003\u8651\u4e86\u6570\u636e\u4fdd\u62a4\u7684\u6cd5\u5f8b\u8981\u6c42\uff0c\u5373\u673a\u5668\u5b66\u4e60\u7b97\u6cd5\u7684\u5206\u6563\u8bad\u7ec3\u4e0d\u9700\u8981\u8de8\u5e73\u53f0\u4ea4\u6362\u6570\u636e\u3002\u56e0\u6b64\uff0c\u8054\u90a6\u5b66\u4e60\u9700\u8981\u4e00\u4e2a\u7edf\u4e00\u7684\u6570\u636e\u6536\u96c6\u548c\u5206\u6790\u6846\u67b6\u624d\u80fd\u53d6\u5f97\u6210\u529f\u3002<\/span><\/section>\n

<\/span><\/section>\n

\n
3.2.2 \u7279\u5f81\u9009\u62e9<\/span><\/strong><\/section>\n

<\/span><\/section>\n<\/h4>\n
\u7279\u5f81\u9009\u62e9\u901a\u5e38\u662f\u7814\u7a76\u4eba\u5458\u6307\u5bfc\u4e0b\u7684\u53d8\u91cf\u9009\u62e9\u3002\u8fd9\u53ef\u4ee5\u57fa\u4e8e\u9886\u57df\u77e5\u8bc6\u6765\u5b9e\u73b0\uff0c\u6216\u8005\u901a\u8fc7\u91c7\u7528\u6700\u5c0f\u5197\u4f59\u6700\u5927\u76f8\u5173\u6027\u7684\u6807\u51c6\u6216\u5176\u4ed6\u7279\u5f81\u9009\u62e9\u6807\u51c6(14)\u6765\u5b9e\u73b0\u6570\u636e\u9a71\u52a8\u3002\u7814\u7a76\u4eba\u5458\u6307\u5bfc\u7684\u7279\u5f81\u9009\u62e9\u53ef\u80fd\u5bf9\u5927\u578b\u6570\u636e\u96c6\u6709\u5e2e\u52a9\uff0c\u4f8b\u5982\u7ecf\u6d4e\u5408\u4f5c\u4e0e\u53d1\u5c55\u7ec4\u7ec7<\/span>\uff08Organisation for Economic Co-operation and Development\u2019s multicountry, OECD\uff09<\/span>\u7684\u591a\u56fd\u8de8\u90e8\u95e8\u56fd\u9645\u5b66\u751f\u8bc4\u4f30\u8ba1\u5212\u6216\u8005\u6765\u81ea\u5065\u5eb7\u548c\u9000\u4f11\u5bb6\u5ead\u7814\u7a76\u7684\u7eb5\u5411\u534f\u8c03\u8001\u9f84\u5316\u8c03\u67e5\u3002<\/span><\/section>\n

<\/span><\/section>\n
\u5728\u9009\u62e9\u8981\u7eb3\u5165\u6a21\u578b\u7684\u7279\u5f81\u65f6\uff0c\u7814\u7a76\u4eba\u5458\u5e94\u8be5\u59cb\u7ec8\u6ce8\u610f\u7ef4\u6570\u707e\u96be\u3002\u7ef4\u6570\u707e\u96be\u63cf\u8ff0\u4e86\u968f\u7740\u95ee\u9898\u7ef4\u5ea6\u7684\u589e\u52a0\uff0c\u6d4b\u8bd5\u8bef\u5dee\u4e5f\u4f1a\u589e\u52a0\u7684\u8d8b\u52bf\uff0c\u9664\u975e\u989d\u5916\u7684\u7279\u5f81\u4e0e\u54cd\u5e94\u771f\u6b63\u76f8\u5173<\/span>\uff08\u5373\u4e0d\u4ec5\u4ec5\u589e\u52a0\u566a\u97f3\uff09<\/span>\u3002\u66f4\u591a\u7684\u7279\u5f81\uff0c\u5373\u6a21\u578b\u4e2d\u7684\u53d8\u91cf\uff0c\u589e\u52a0\u4e86\u95ee\u9898\u7684\u7ef4\u6570\uff0c\u52a0\u5267\u4e86\u8fc7\u62df\u5408\u7684\u98ce\u9669\u3002\u56e0\u6b64\uff0c\u53ef\u4ee5\u6536\u96c6\u6570\u5343\u751a\u81f3\u6570\u767e\u4e07\u7279\u5f81\u7684\u6570\u636e\u91c7\u96c6\u65b9\u9762\u7684\u8fdb\u5c55\u662f\u4e00\u628a\u53cc\u5203\u5251\u3002\u5982\u679c\u7279\u5f81\u771f\u6b63\u76f8\u5173\uff0c\u6837\u672c\u5177\u6709\u4ee3\u8868\u6027\uff0c\u90a3\u4e48\u53ef\u4ee5\u63d0\u5347\u9884\u6d4b\u6548\u679c\u3002\u4f46\u5982\u679c\u4e0d\u662f\uff0c\u5c06\u4f1a\u5bfc\u81f4\u66f4\u52a0\u6709\u504f\u7684\u7ed3\u679c\u3002\u6b64\u5916\uff0c\u5373\u4f7f\u5b83\u4eec\u662f\u76f8\u5173\u7684\uff0c\u504f\u5dee\u7684\u51cf\u5c11\u53ef\u80fd\u4f1a\u88ab\u589e\u52a0\u7684\u65b9\u5dee\u62b5\u6d88(5)\u3002\u7a00\u758f\u6062\u590d\u7684 Donoho-Tanner \u76f8\u53d8\u5b9a\u4e49\u4e86\u7a00\u758f\/\u6b20\u91c7\u6837\u6743\u8861\u66f2\u7ebf\u7684\u5c16\u9510\u8fb9\u754c(13) ; \u7279\u522b\u662f\uff0c\u5f53\u6a21\u578b\u7684\u590d\u6742\u6027\u589e\u52a0\u8d85\u8fc7\u9608\u503c\u65f6\uff0c\u6a21\u578b\u9009\u62e9\u548c\u62df\u5408\u7a81\u7136\u4e2d\u65ad\uff0c\u8fd9\u53d6\u51b3\u4e8e\u6570\u636e\u5927\u5c0f\u548c\u771f\u5b9e\u590d\u6742\u5ea6\u3002\u5bf9\u4e8e\u53d8\u91cf\u6216\u7279\u5f81\u9009\u62e9\uff0c\u8fd9\u9650\u5236\u4e86\u6570\u636e\u5206\u6790\u7684\u6210\u529f\u7a0b\u5ea6\u3002\u6570\u636e\u9a71\u52a8\u7684\u7279\u5f81\u9009\u62e9\u662f\u60e9\u7f5a\u65b9\u6cd5\u7684\u7279\u6027\uff0c\u60e9\u7f5a\u65b9\u6cd5\u6839\u636e\u7279\u5f81\u7684\u9884\u6d4b\u80fd\u529b\u6765\u9009\u62e9\u7279\u5f81\uff0c\u4ece\u800c\u9650\u5236\u4e86\u6a21\u578b\u4e2d\u7279\u5f81\u7684\u6570\u91cf\u3002<\/span><\/section>\n

<\/span><\/section>\n

\n
3.2.3 \u7279\u5f81\u5de5\u7a0b<\/span><\/strong><\/section>\n

<\/span><\/section>\n<\/h4>\n
\u7406\u8bba\u4e0a\u6709\u8da3\u7684\u7279\u5f81\uff0c\u4f8b\u5982\uff0c\u7d2f\u79ef\u98ce\u9669<\/span>\uff08\u4e24\u4e2a\u9884\u6d4b\u56e0\u5b50\u7684\u4e58\u6570\u6548\u5e94\uff09<\/span>\u6216\u6d4b\u91cf\u4e4b\u95f4\u7684\u53d8\u5316<\/span>\uff08\u4f8b\u5982\u968f\u7740\u65f6\u95f4\u7684\u63a8\u79fb\u4f53\u91cd\u51cf\u8f7b\uff09<\/span>\uff0c\u4e0d\u80fd\u7b80\u5355\u5730\u6dfb\u52a0\u5230\u53d8\u91cf\u6c60\u4e2d\u4ee5\u5b8c\u6210\u68c0\u6d4b\u4efb\u52a1\u3002\u9700\u8981\u6839\u636e\u9886\u57df\u77e5\u8bc6\u8bbe\u8ba1\u7279\u5f81\uff0c\u4f8b\u5982\uff0c\u6784\u5efa\u5dee\u5f02\u5ea6\u91cf<\/span>\uff08difference measures\uff09<\/span>\u6216\u5e73\u65b9\u9879<\/span>\uff08squared terms\uff09<\/span>\u3002\u6700\u8fd1\u63d0\u51fa\u4e86\u4e00\u79cd\u7814\u7a76\u7cfb\u7edf\u7279\u5f81\u4ea4\u4e92\u7684\u65b9\u6cd5\uff0c\u5373\u57fa\u4e8e\u6811\u7684\u968f\u673a\u68ee\u6797\u7279\u5f81\u91cd\u8981\u6027\u548c\u7279\u5f81\u4ea4\u4e92\u7f51\u7edc\u5206\u6790\u6846\u67b6\uff0815\uff09\u3002\u672c\u6587\u8fd8\u5efa\u8bae\u63a2\u8ba8<\/span>\uff08\u76f8\u5173\uff09<\/span>\u81ea\u53d8\u91cf\u96c6\u5408\u4e2d\u590d\u6742\u6027\u7684\u964d\u4f4e\u662f\u5426\u6709\u610f\u4e49\uff0c\u4f8b\u5982\u57fa\u4e8e\u7406\u8bba\u7684\u56e0\u7d20\u5206\u6790\u6216\u6570\u636e\u805a\u7c7b\u6216\u53d8\u91cf\u9009\u62e9\uff0c\u4ee5\u63d0\u9ad8\u7279\u5f81\u7684\u6bd4\u4f8b\uff0c\u5e76\u4e14\u9488\u5bf9\u5177\u4f53\u7684\u7814\u7a76\u9886\u57df\u548c\u65b9\u6cd5\u89c4\u8303\u63d0\u51fa\u4e86\u6837\u672c\u89c4\u6a21\u7684\u5efa\u8bae(16)\u3002\u5982\u679c\u6570\u636e\u96c6\u8db3\u591f\u5927\uff0c\u4e00\u4e2a\u5e38\u89c4\u7ecf\u9a8c\u5c31\u662f\u6570\u4e07\u4e2a\u76f8\u5173\u5355\u5143<\/span>\uff08\u4f8b\u5982\uff0c\u4e00\u4e2a\u8c03\u67e5\u7684\u53d7\u8bbf\u8005\uff09<\/span>\u5373\u53ef\u7528\uff1b\u8fd9\u4f7f\u5f97\u795e\u7ecf\u7f51\u7edc\u7684\u4f7f\u7528\u6210\u4e3a\u53ef\u80fd\uff0c\u4f17\u6240\u5468\u77e5\uff0c\u795e\u7ecf\u7f51\u7edc\u4f7f\u7528\u7279\u5f81\u5de5\u7a0b\u6765\u751f\u6210\u6a21\u578b\u3002<\/span><\/section>\n

<\/span><\/section>\n
\u867d\u7136\u6709\u5fc5\u8981\u4fdd\u8bc1\u8fde\u7eed\u53d8\u91cf\u5177\u6709\u7b49\u6548\u65b9\u5dee\uff0c\u4f46\u4f7f\u7528\u624b\u5de5\u7279\u5f81\u5de5\u7a0b\u5c40\u9650\u4e8e\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u7684\u7814\u7a76\u4eba\u5458\u786e\u4fdd\u73b0\u5b9e\u4e16\u754c\u5e94\u7528\u4e2d\u7684\u4e00\u4e9b\u53ef\u89e3\u91ca\u6027\uff0c\u7279\u522b\u662f\u5728\u9700\u8981\u660e\u786e\u73b0\u8c61\u5b9a\u4e49\u4ee5\u4fbf\u5728\u73b0\u5b9e\u4e16\u754c\u5e72\u9884\u7684\u60c5\u51b5\u4e0b\u8fdb\u884c\u56e0\u679c\u63a8\u65ad(17)\u3002\u5047\u8bbe\u6211\u4eec\u53cd\u590d\u8bc4\u4f30\u8eab\u4f53\u8d28\u91cf\u6307\u6570<\/span>\uff08Body mass index, BMI\uff09<\/span>\u6765\u9884\u6d4b\u6216\u89e3\u91ca\u5065\u5eb7\u72b6\u51b5\uff0c\u8bc4\u4f30\u65f6\u6807\u8bb0 BMI \u7684\u6a21\u578b\u4e3a No.x<\/span>\uff08\u6216\u75be\u75c5\u53d1\u4f5c\u524d x \u5e74\uff09<\/span>\uff0c\u8fd9\u5728\u5b9e\u8df5\u4e2d\u53ef\u80fd\u6ca1\u6709\u610f\u4e49; \u7136\u800c\uff0c\u5728\u624b\u52a8\u7279\u5f81\u5de5\u7a0b\u540e\uff0c\u4f5c\u4e3a\u75be\u75c5\u9884\u6d4b\u7684\u6a21\u578b\u6807\u8bb0BMI\u7684\u589e\u52a0\u6216\u51cf\u5c11\u53ef\u80fd\u5bf9\u8bc6\u522b\u9ad8\u5371\u60a3\u8005\u975e\u5e38\u6709\u7528\u3002\u53e6\u4e00\u65b9\u9762\uff0c\u4f20\u7edf\u7684\u5efa\u6a21\u9ad8\u5ea6\u4f9d\u8d56\u4e8e\u7814\u7a76\u4eba\u5458\u7684\u51b3\u7b56<\/span>\uff08\u4f8b\u5982\uff0c\u5efa\u7acb\u4e8c\u6b21\u5173\u7cfb\u800c\u4e0d\u662f\u7ebf\u6027\u5173\u7cfb\u7684\u6a21\u578b\u4ee5\u53ca\u624b\u52a8\u5efa\u7acb\u4ea4\u4e92\u6548\u5e94\u7684\u6a21\u578b\uff09<\/span>\uff0c\u5982\u795e\u7ecf\u7f51\u7edc\u7684\u57fa\u4e8e\u7b97\u6cd5\u7684\u51b3\u7b56\u5173\u4e8e\u7279\u5f81\u5de5\u7a0b\u53ef\u4ee5\u63d0\u4f9b\u66f4\u7a33\u5065\u548c\u51c6\u786e\u7684\u53d1\u73b0\u3002<\/span><\/section>\n

<\/span><\/section>\n

\n
3.3 \u6a21\u578b\u5efa\u7acb<\/span><\/strong><\/span><\/section>\n<\/h3>\n

<\/section>\n
\u5728\u5efa\u6a21\u8fc7\u7a0b\u4e2d\uff0c\u673a\u5668\u5b66\u4e60\u65b9\u6cd5\u901a\u5e38\u5305\u62ec\u4e09\u4e2a\u6b65\u9aa4: (i)\u8bad\u7ec3\u3001(ii)\u9a8c\u8bc1\u548c(iii)\u6d4b\u8bd5\u3002\u5927\u591a\u6570\u7814\u7a76\u4eba\u5458\u90fd\u719f\u6089\u673a\u5668\u5b66\u4e60\u5efa\u6a21\u8fc7\u7a0b\uff0c\u5176\u4e2d\u5305\u62ec\u9996\u5148\u5c06\u6570\u636e\u5206\u5272\u6210\u4e00\u4e2a\u8bad\u7ec3\u548c\u4e00\u4e2a\u72ec\u7acb\u7684\u6d4b\u8bd5\u96c6\uff0c\u7136\u540e\u8fdb\u4e00\u6b65\u5c06\u8bad\u7ec3\u6570\u636e\u5206\u5272\u6210\u8bad\u7ec3\u96c6\u548c\u9a8c\u8bc1\u96c6\u3002<\/span><\/section>\n

<\/span><\/section>\n
\uff081\uff09\u8bad\u7ec3\uff1a\u5728\u8bad\u7ec3\u96c6\u4e2d\u4f30\u8ba1\u6a21\u578b\u53c2\u6570\u3002<\/span><\/section>\n

<\/span><\/section>\n
\uff082\uff09\u9a8c\u8bc1\uff1a\u5229\u7528\u9a8c\u8bc1\u96c6\u8bc4\u4f30\u5df2\u8bad\u7ec3\u7684\u6a21\u578b\u4ee5\u9009\u62e9\u6700\u63a5\u8fd1\u76f8\u5173\u6307\u6807\uff08\u5982\u590d\u6742\u5ea6\uff09\u7684\u6a21\u578b\uff0c\u5e76\u8c03\u6574\u6700\u9002\u5408\u7684\u8d85\u53c2\u6570\u3002<\/span><\/section>\n

<\/section>\n
\uff083\uff09\u6d4b\u8bd5\uff1a\u8be5\u6a21\u578b\u5728\u4e00\u4e2a\u5355\u72ec\u7684<\/span>\uff08\u4fdd\u7559\u7684\uff09<\/span>\u6d4b\u8bd5\u6570\u636e\u96c6\u4e0a\u8fdb\u884c\u6d4b\u8bd5\uff0c\u4ee5\u8bc4\u4f30\u5176\u6cdb\u5316\u8bef\u5dee\u3002\u8fd9\u4e2a\u6d4b\u91cf\u7ed3\u679c\u8868\u660e\u4e86\u8be5\u6a21\u578b\u5728\u672a\u6765\u7684\u6570\u636e\u96c6\u4e0a\u7684\u6027\u80fd\u8868\u73b0\uff0c\u4f8b\u5982\u5206\u7c7b\u7684\u51c6\u786e\u6027\u3002<\/span><\/section>\n

<\/span><\/section>\n
\u4e3a\u4e86\u63d0\u5347\u9a8c\u8bc1\u96c6\u548c\u6d4b\u8bd5\u96c6\u4e0a\u7684\u6027\u80fd\uff0c\u53ef\u4ee5\u91c7\u7528\u4ea4\u53c9\u9a8c\u8bc1\u7684\u65b9\u6cd5\u3002\u8be5\u65b9\u6cd5\u540c\u6837\u88ab\u5e94\u7528\u4e8e\u4f20\u7edf\u7684\u7edf\u8ba1\u65b9\u6cd5\u4e2d\u3002<\/span><\/section>\n

<\/span><\/section>\n
\u7814\u7a76\u4eba\u5458\u9700\u8981\u6ce8\u610f\u9ad8\u51c6\u786e\u7387\u4e0e\u8fc7\u62df\u5408\u4e4b\u95f4\u7684\u6743\u8861\u3002\u7531\u4e8e\u8fc7\u62df\u5408\uff0c\u8bad\u7ec3\u8bef\u5dee\u901a\u5e38\u968f\u6a21\u578b\u590d\u6742\u5ea6\u7684\u589e\u52a0\u800c\u51cf\u5c0f\u3002\u6d4b\u8bd5\u8bef\u5dee\u66f2\u7ebf\u5448 U \u5f62\uff0c\u5148\u7531\u4e8e\u8fc7\u62df\u5408\u800c\u51cf\u5c0f\uff0c\u800c\u540e\u7531\u4e8e\u8fc7\u62df\u5408\u800c\u589e\u5927<\/span>\uff08\u5982\u56fe1\u6240\u793a\uff09<\/span>\u3002<\/span><\/section>\n

<\/span><\/section>\n
<\/section>\n
\u6a21\u578b\u8bef\u5dee\u548c\u590d\u6742\u5ea6\u4e4b\u524d\u6743\u8861\u7684\u793a\u610f\u56fe<\/span><\/section>\n

<\/span><\/section>\n
\u867d\u7136\u8fd9\u79cd\u60c5\u51b5\u5177\u6709\u6700\u5c0f\u6d4b\u8bd5\u8bef\u5dee\uff0c\u4f46\u4e00\u5b9a\u4e0d\u8981\u4f7f\u7528\u6d4b\u8bd5\u6570\u636e\u8fdb\u884c\u6a21\u578b\u62df\u5408\u6216\u9009\u62e9\u6a21\u578b\uff0c\u5426\u5219\u5c06\u5f97\u5230\u4e00\u4e2a\u8fc7\u4e8e\u4e50\u89c2\u4e14\u4e0d\u5207\u5b9e\u9645\u7684\u6027\u80fd\u6d4b\u91cf\u503c\u3002\u56e0\u6b64\uff0c\u5728\u9a8c\u8bc1\u9636\u6bb5\uff0c\u8bad\u7ec3\u6570\u636e\u66f4\u8fdb\u4e00\u6b65\u5730\u88ab\u5212\u5206\u4e3a\u9a8c\u8bc1\u96c6\uff0c\u5e76\u4e14\u6a21\u578b\u590d\u6742\u5ea6\u548c\u8c03\u53c2\u662f\u5728\u6700\u5c0f\u5316\u9a8c\u8bc1\u96c6\u8bef\u5dee\u65f6\u5b8c\u6210\u7684\u3002\u4e3a\u4e86\u63d0\u5347\u4f30\u8ba1\u51c6\u786e\u5ea6\u5e76\u964d\u4f4e\u65b9\u5dee\uff0c\u901a\u5e38\u5728K\u6298\u6570\u636e\u96c6\u4e0a\u8fdb\u884c\u4ea4\u53c9\u9a8c\u8bc1\u3002\u7136\u800c\uff0c\u5728\u6d4b\u8bd5\u4e2d\u4e5f\u53ef\u4ee5\u8fdb\u884c\u4ea4\u53c9\u9a8c\u8bc1\u4ee5\u6539\u5584\u5bf9\u6d4b\u8bd5\u8bef\u5dee\u7684\u4f30\u8ba1\uff0c\u7279\u522b\u662f\u5728\u6837\u672c\u91cf\u8f83\u5c0f\u7684\u60c5\u51b5\u4e0b\u3002\u4e3a\u4e86\u786e\u4fdd\u6d4b\u8bd5\u6570\u636e\u6ca1\u6709\u7ecf\u8fc7\u8bad\u7ec3\uff0c\u53ef\u80fd\u9700\u8981\u5d4c\u5957K\u6b21\u7684\u64cd\u4f5c\uff0c\u5373\u4f7f\u7528K\u6298\u9a8c\u8bc1\u6765\u9009\u62e9\u6bcf\u4e2a\u6d4b\u8bd5\u5185\u7684\u6a21\u578b\u3002\u6b64\u5916\uff0c\u5982\u679c\u4f7f\u7528\u4ea4\u53c9\u9a8c\u8bc1\u5f97\u5230\u7684\u6d4b\u8bd5\u8bef\u5dee\uff0c\u5219\u5e94\u8be5\u6307\u660e\u968f\u673a\u5212\u5206\u7684\u91cd\u590d\u6027\uff0c\u4ee5\u5141\u8bb8\u5176\u4ed6\u7814\u7a76\u4eba\u5458\u8fdb\u884c\u6bd4\u8f83\u548c\u6d4b\u8bd5\u3002\u56fe1\u8bf4\u660e\u4e86\u9884\u6d4b\u51c6\u786e\u6027\u548c\u8fc7\u62df\u5408\u4e4b\u95f4\u7684\u6743\u8861\uff0c\u7a81\u51fa\u4e86\u8bad\u7ec3\u3001\u9a8c\u8bc1\u548c\u6d4b\u8bd5\u8bef\u5dee\u4e4b\u95f4\u7684\u76f8\u4e92\u5173\u7cfb\u3002\u6211\u4eec\u603b\u662f\u671f\u671b\u8bad\u7ec3\u8bef\u5dee\u66f2\u7ebf\u4f4e\u4e8e\u6d4b\u8bd5\u8bef\u5dee\u66f2\u7ebf\uff0c\u8fd9\u662f\u56e0\u4e3a\u5927\u591a\u6570\u65b9\u6cd5\u7684\u76ee\u6807\u662f\u6700\u5c0f\u5316\u8bad\u7ec3\u8bef\u5dee\u3002\u9a8c\u8bc1\u8bef\u5dee\u66f2\u7ebf\u901a\u5e38\u4f4d\u4e8e\u6d4b\u8bd5\u8bef\u5dee\u66f2\u7ebf\u4e4b\u4e0a\uff0c\u56e0\u4e3a\u5b83\u662f\u5728\u8f83\u5c0f\u7684\u8bad\u7ec3\u96c6\u4e0a\u8fdb\u884c\u8bad\u7ec3\u7684\uff0c\u800c\u4e14\u7531\u4e8e\u6570\u636e\u968f\u673a\u5212\u5206\uff0c\u9a8c\u8bc1\u8bef\u5dee\u66f2\u7ebf\u4e5f\u53ef\u80fd\u5177\u6709\u9ad8\u5ea6\u7684\u53ef\u53d8\u6027\uff0c\u5c3d\u7ba1\u4ea4\u53c9\u9a8c\u8bc1\u6709\u52a9\u4e8e\u51cf\u5c11\u8fd9\u79cd\u53ef\u53d8\u6027\u3002\u7136\u800c\uff0c\u9a8c\u8bc1\u7684\u76ee\u6807\u662f\u786e\u5b9a\u6a21\u578b\u7684\u7075\u6d3b\u6027\uff0c\u5373\u6d4b\u8bd5\u8bef\u5dee\u7684\u6700\u5c0f\u70b9\u3002\u6ce8\u610f\u5230\uff0c\u5728\u5047\u8bbe\u6570\u636e\u662f\u72ec\u7acb\u4e14\u5206\u5e03\u76f8\u540c\u7684\u60c5\u51b5\u4e0b\uff0c\u4ea4\u53c9\u9a8c\u8bc1\u662f\u5408\u9002\u7684\uff0c\u4f46\u5bf9\u4e8e\u67d0\u4e9b\u6570\u636e\uff0c\u4f8b\u5982\u65f6\u95f4\u5e8f\u5217\u6216\u7eb5\u5411\u6570\u636e\uff0c\u8fd9\u662f\u4e0d\u5408\u9002\u7684\uff0c\u5fc5\u987b\u8003\u8651\u5230\u6570\u636e\u7684\u7ed3\u6784\u3002\u5f53\u9a8c\u8bc1\u66f2\u7ebf\u76f8\u5bf9\u5e73\u5766\u65f6\uff0c\u6700\u597d\u4f7f\u7528\u66f4\u7b80\u5355\u7684\u6a21\u578b\u3002<\/span><\/section>\n

<\/span><\/section>\n
\u5728\u6a21\u578b\u5efa\u7acb\u9636\u6bb5\uff0c\u8bb8\u591a\u9884\u6d4b\u673a\u5668\u5b66\u4e60\u65b9\u6cd5\u4f1a\u7ed9\u7279\u5f81<\/span>\uff08\u53d8\u91cf\uff09<\/span>\u5206\u914d\u91cd\u8981\u6027\uff0c\u4ee5\u66f4\u5b9e\u8d28\u6027\u5730\u63d0\u9ad8\u6027\u80fd\u3002\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u7684\u7814\u7a76\u4eba\u5458\u9700\u8981\u610f\u8bc6\u5230\uff0c\u4e00\u4e9b\u673a\u5668\u5b66\u4e60\u65b9\u6cd5\uff0c\u4f8b\u5982\u56de\u5f52\u6811\uff0c\u4ec5\u4ec5\u56e0\u4e3a\u6709\u66f4\u591a\u53ef\u80fd\u7684\u5206\u88c2\u70b9\uff0c\u5c31\u4f1a\u9ad8\u4f30\u8fde\u7eed\u9884\u6d4b\u7684\u503c\u3002\u867d\u7136\u5728\u67d0\u4e9b\u7b97\u6cd5\u4e2d\u5b58\u5728\u5206\u5272\u5206\u7c7b\u53d8\u91cf\u51b3\u7b56\u7684\u89c4\u5219\uff0c\u4f46\u5b83\u4eec\u5728\u4e0d\u540c\u8f6f\u4ef6\u5305\u4e4b\u95f4\u7684\u5904\u7406\u65b9\u5f0f\u662f\u4e0d\u540c\u7684\u3002\u76f8\u6bd4\u4e4b\u4e0b\uff0c\u5982BART<\/span>\uff08\u5728\u4e0b\u9762\u66f4\u8be6\u7ec6\u5730\u63cf\u8ff0\uff09<\/span>\u8fd9\u6837\u7684\u65b9\u6cd5\u5e94\u8be5\u540c\u65f6\u5904\u7406\u8fde\u7eed\u3001\u4e8c\u5206\u548c\u5206\u7c7b\u9884\u6d4b\u7684\u95ee\u9898\u3002\u7ecf\u9a8c\u8868\u660e\uff0c\u8fde\u7eed\u53d8\u91cf\u53ef\u80fd\u4ecd\u7136\u53d7\u5230\u9752\u7750<\/span>\u3002<\/span><\/section>\n

<\/span><\/section>\n
\u6a21\u578b\u5efa\u7acb\u53ef\u4ee5\u901a\u8fc7\u53cb\u597d\u7684\u7528\u6237\u754c\u9762\u6765\u5b8c\u6210\uff0c\u8fd9\u4e9b\u754c\u9762\u662f\u5728\u8fc7\u53bb\u5341\u5e74\u4e2d\u5f00\u53d1\u7684\u3002AutoML \u65b9\u6cd5\u65e8\u5728\u4fc3\u8fdb\u975e\u6570\u636e\u79d1\u5b66\u7814\u7a76\u4eba\u5458\u5e94\u7528\u673a\u5668\u5b66\u4e60(19)\u3002\u9ed8\u8ba4\u60c5\u51b5\u4e0b\uff0cAutoML \u5bf9\u5e7f\u4e49\u7ebf\u6027\u6a21\u578b\u3001\u68af\u5ea6\u63d0\u5347\u673aGradient Boosting Machine\u3001\u968f\u673a\u68ee\u6797\u548c\u6df1\u5ea6\u795e\u7ecf\u7f51\u7edc\u8fdb\u884c\u8bad\u7ec3\u548c\u4ea4\u53c9\u9a8c\u8bc1\uff0c\u5e76\u901a\u8fc7\u8d85\u5b66\u4e60<\/span>\uff08SuperLearning\uff09<\/span>\u6216\u96c6\u6210\u7ed3\u5408\u6a21\u578b\u4ee5\u6539\u8fdb\u62df\u5408\u6548\u679c\u3002Python\u7684\u8f6f\u4ef6\u5e93scikit-learn \u5141\u8bb8\u7814\u7a76\u4eba\u5458\u901a\u8fc7\u4e00\u4e2a\u7528\u6237\u53cb\u597d\u7684\u5e73\u53f0\u6765\u9009\u62e9\u673a\u5668\u5b66\u4e60\u7b97\u6cd5(20)\u3002\u867d\u7136\u6613\u4e8e\u4f7f\u7528\u7684\u754c\u9762\u5bf9\u975e\u6570\u636e\u79d1\u5b66\u7814\u7a76\u4eba\u5458\u5177\u6709\u5438\u5f15\u529b\uff0c\u5e76\u4e14\u5728\u8bb8\u591a\u60c5\u51b5\u4e0b\u53ef\u80fd\u6709\u5e2e\u52a9\uff0c\u4f46\u662f\u6211\u4eec\u4ecd\u7136\u8ba4\u4e3a\uff0c\u4e0e\u4f20\u7edf\u63a8\u65ad\u7edf\u8ba1\u5b66\u7684\u7814\u7a76\u7c7b\u4f3c\uff0c\u5bf9\u5e94\u7528\u65b9\u6cd5\u66f4\u8be6\u7ec6\u7684\u7406\u89e3\u5bf9\u5b9e\u8d28\u6027\u7684\u7814\u7a76\u662f\u6709\u8d21\u732e\u7684\u3002<\/span><\/section>\n

<\/span><\/section>\n

\n
3.4 \u771f\u5b9e\u4e16\u754c\u5e94\u7528\u673a\u5668\u5b66\u4e60\u7684\u8003\u8651<\/span><\/strong><\/span><\/section>\n

<\/span><\/section>\n<\/h3>\n
\u5728\u673a\u5668\u5b66\u4e60\u57fa\u7840\u4e0a\u5f00\u53d1\u7684\u6a21\u578b\uff0c\u7c7b\u4f3c\u4e8e\u66f4\u201c\u4f20\u7edf\u201d\u7684\u6a21\u578b\uff0c\u80fd\u591f\u5728\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u7684\u653f\u7b56\u548c\u5b9e\u8df5\u4e2d\u544a\u77e5\u9884\u9632\u3001\u8bca\u65ad\u3001\u6cbb\u7597\u548c\u62a4\u7406\u3002\u5229\u76ca\u76f8\u5173\u8005\u3001\u7528\u6237\u4ee5\u53ca\u6574\u4e2a\u793e\u4f1a\u90fd\u9700\u8981\u63a5\u53d7\u673a\u5668\u5b66\u4e60\u5728\u73b0\u5b9e\u4e16\u754c\u4e2d\u7684\u5e94\u7528\u3002\u7b80\u800c\u8a00\u4e4b\uff0c\u673a\u5668\u5b66\u4e60<\/span>\uff08\u5728\u8fd9\u79cd\u60c5\u51b5\u4e0b\u901a\u5e38\u88ab\u79f0\u4e3a\u4eba\u5de5\u667a\u80fd\uff09<\/span>\u9700\u8981\u201c\u503c\u5f97\u4fe1\u8d56\u201d\u3002\u7279\u522b\u76f8\u5173\u7684\u662f\uff0c\u5f53\u4e3a\u673a\u5668\u5b66\u4e60\u9a71\u52a8\u7684\u533b\u7597\u4fdd\u5065\u6216\u793e\u4f1a\u5e94\u7528\u5efa\u7acb\u66f4\u5927\u7684\u6846\u67b6\u65f6\uff0c\u9700\u8981\u4fdd\u8bc1\u673a\u5668\u5b66\u4e60\u548c\u4eba\u5de5\u667a\u80fd\u5728\u6280\u672f\u548c\u793e\u4f1a\u7684\u89d2\u5ea6\u90fd\u662f\u5408\u6cd5\u7684\u3001\u9053\u5fb7\u7684\u4ee5\u53ca\u5f3a\u5927\u7684(21)\u3002\u6211\u4eec\u5efa\u8bae\u8bfb\u8005\u53c2\u8003\u6700\u8fd1\u51fa\u7248\u7269\u4e2d\u5173\u4e8e\u201c\u53ef\u4fe1\u8d56\u7684\u4eba\u5de5\u667a\u80fd\u201d\u7684\u8ba8\u8bba(21)\u3002\u5728\u63a5\u4e0b\u6765\u7684\u6587\u7ae0\u4e2d\uff0c\u5f53\u6211\u4eec\u5c06\u673a\u5668\u5b66\u4e60\u5e94\u7528\u4e8e\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u65f6\uff0c\u6211\u4eec\u5c06\u53c2\u8003\u4e09\u4e2a\u76f8\u5173\u7684\u53ef\u4fe1\u8d56\u4eba\u5de5\u667a\u80fd\u7684\u6982\u5ff5\uff0c\u7279\u522b\u662f(i)\u53ef\u89e3\u91ca\u6027\u3001(ii)\u516c\u5e73\u6027\u3001(iii)\u666e\u904d\u6027\u4ee5\u53ca(iv)\u673a\u5668\u5b66\u4e60\u6765\u8f85\u52a9\u4eba\u7c7b\u7684\u80fd\u529b\u548c\u6280\u80fd\u3002<\/span><\/section>\n

<\/span><\/section>\n

\n
3.4.1 \u53ef\u89e3\u91ca\u6027\u4e0e\u53ef\u89c6\u5316<\/span><\/strong><\/section>\n

<\/span><\/section>\n<\/h4>\n
\u53ef\u89e3\u91ca\u7684\u673a\u5668\u5b66\u4e60\u610f\u5473\u7740\u4ece\u673a\u5668\u5b66\u4e60\u6a21\u578b\u4e2d\u63d0\u53d6\u76f8\u5173\u77e5\u8bc6\uff0c\u4f8b\u5982\uff0c\u80fd\u591f\u8f93\u5165\u5b66\u79d1\u9886\u57df\u77e5\u8bc6\uff0c\u5e76\u62e5\u6709\u9884\u6d4b\u51c6\u786e\u5ea6\u3001\u63cf\u8ff0\u51c6\u786e\u6027\u548c\u76f8\u5173\u6027\u7b49\u5c5e\u6027\uff0c\u5176\u4e2d\u76f8\u5173\u6027\u662f\u76f8\u5bf9\u4e8e\u4eba\u7c7b\u53d7\u4f17\u8fdb\u884c\u5224\u65ad\u7684(9)\u3002\u53ef\u89e3\u91ca\u6027\u5728\u9ad8\u98ce\u9669\u51b3\u7b56\u4e2d\u5c24\u4e3a\u91cd\u8981\uff0c\u4f8b\u5982\u4e34\u5e8a\u51b3\u7b56\u3002\u4e0d\u540c\u4e8e\u5bfb\u627e\u539f\u59cb\u201c\u9ed1\u76d2\u201d\u6a21\u578b\u7684\u53ef\u89e3\u91ca\u6027\uff0c\u90e8\u5206\u5b66\u8005\u63d0\u51fa\u9700\u8981\u8bbe\u8ba1\u53ef\u89e3\u91ca\u6a21\u578b\uff0c\u56e0\u4e3a\u53ea\u6709\u53ef\u89e3\u91ca\u6a21\u578b\u624d\u80fd\u5b9e\u73b0\u7528\u6237\u773c\u4e2d\u7684\u900f\u660e\u5ea6\u548c\u53ef\u4fe1\u5ea6(22)\u3002\u53ef\u89e3\u91ca\u7684\u4eba\u5de5\u667a\u80fd\u5728\u66f4\u8fd1\u671f\u7684\u9879\u76ee\u4e2d\u901a\u5e38\u88ab\u79f0\u4e3a\u53ef\u89e3\u91ca\u7684\u4eba\u5de5\u667a\u80fd<\/span>\uff08explainable AI, XAI\uff09<\/span>(23)\u3002<\/span><\/section>\n

<\/span><\/section>\n
\u5728\u5b9e\u8df5\u4e2d\uff0c\u53ef\u89e3\u91ca\u6027\u5e38\u5e38\u4e0e\u53ef\u89c6\u5316\u53d8\u91cf\u4e4b\u95f4\u9884\u4f30\u6216\u627e\u5230\u7684\u5173\u7cfb\u6709\u5173\u3002\u4e3a\u4e86\u63d0\u9ad8\u53ef\u89e3\u91ca\u6027\uff0c\u901a\u8fc7\u6f5c\u5728\u7684\u975e\u7ebf\u6027\u6a21\u5f0f\u6216\u975e\u968f\u673a\u7f3a\u5931\u7684\u6570\u636e\uff0c\u53ef\u4ee5\u53ef\u89c6\u5316\u4e8c\u53d8\u91cf\u548c\u9ad8\u9636\u6570\u636e\u7684\u5173\u8054\uff0c\u4f8b\u5982\u90e8\u5206\u4f9d\u8d56\u56fe<\/span>\uff08partial dependence plots \uff09<\/span>(24)\u3001SHAP\u56fe<\/span>\uff08Shapley Additive exPlanations, SHAP\uff09<\/span>(25,26)\u6216\u4e2a\u4f53\u6761\u4ef6\u671f\u671b\u56fe<\/span>\uff08individual conditional expectation plots\uff09<\/span>\uff0c\u4e5f\u88ab\u79f0\u4e3a\u9002\u5e94\u90e8\u5206\u4f9d\u8d56\u6027\u56fe<\/span>\uff08adaptation of partial dependence plots\uff09<\/span>(27)\u3002<\/span><\/section>\n

<\/span><\/section>\n

\n
3.4.2 \u516c\u5e73\u6027<\/span><\/strong><\/section>\n

<\/span><\/section>\n<\/h4>\n
\u5728\u5065\u5eb7\u548c\u793e\u4f1a\u79d1\u5b66\u4e2d\uff0c\u4e00\u4e2a\u6982\u5ff5\u662f\u673a\u5668\u5b66\u4e60\u7b97\u6cd5\u8981\u201c\u516c\u5e73\u201d\u5730\u51b3\u5b9a\uff0c\u5373\u4e0d\u6b67\u89c6\u67d0\u4e2a\u793e\u4f1a\u6216\u5c11\u6570\u7fa4\u4f53\u3002\u672c\u6587\u53c2\u8003\u4e86\u5176\u4ed6\u6587\u732e\u4e2d\u4e0eML\u4e2d\u7684\u516c\u5e73\u6027\u6709\u5173\u7684\u4f8b\u5b50(28)\u4ee5\u53ca\u4e00\u4e2a\u536b\u751f\u4fdd\u5065\u7b97\u6cd5\u4e2d\u79cd\u65cf\u504f\u89c1\u7684\u7a81\u51fa\u4f8b\u5b50(29)\u3002\u5728\u8fd9\u91cc\uff0c\u7279\u522b\u662f\u5728\u673a\u5668\u5b66\u4e60\u5728\u9ad8\u98ce\u9669\u51b3\u7b56<\/span>\uff08\u5982\u9884\u6d4b\u7d2f\u72af\uff09<\/span>\u4e2d\u7684\u5e94\u7528\u4e2d\uff0c\u9700\u8981\u5e73\u8861\u6a21\u578b\u7684\u9884\u6d4b\u51c6\u786e\u6027\u4e0e\u6240\u6709\u793e\u4f1a\u7fa4\u4f53\u516c\u5e73\u6027\uff0c\u5373\u673a\u5668\u5b66\u4e60\u6a21\u578b\u4e2d\u6240\u6709\u793e\u4f1a\u7fa4\u4f53\u83b7\u5f97\u7406\u60f3\u7ed3\u679c\u7684\u6982\u7387\u76f8\u7b49\u3002\u91cd\u8981\u7684\u662f\uff0c\u7ed3\u6784\u6027\u6b67\u89c6\/\u79cd\u65cf\u4e3b\u4e49\u4f1a\u65e0\u610f\u4e2d\u7834\u574f\u673a\u5668\u5b66\u4e60\u51b3\u7b56\u4e2d\u7684\u5065\u5eb7\u516c\u5e73\u3002\u8fd9\u79cd\u201c\u7b97\u6cd5\u516c\u5e73\u201d\u7684\u6982\u5ff5\u53d7\u5230\u4e86\u6279\u8bc4\u3002\u89e3\u51b3\u201c\u5728\u6570\u636e\u4e2d\u201d\u4ea7\u751f\u516c\u5e73\u95ee\u9898\u7684\u65b0\u65b9\u6cd5\u5efa\u8bae\u5b9a\u4e49\u53cd\u4e8b\u5b9e\u516c\u5e73(30)\u3002\u6700\u8fd1\u7684\u4e00\u9879\u7cfb\u7edf\u8bc4\u4f30\u8868\u660e\uff0c\u4f7f\u7528\u516c\u5e73\u6807\u51c6\u4f1a\u635f\u5bb3\u6a21\u578b\u7684\u6027\u80fd\uff0c\u540c\u65f6\u4e5f\u4f1a\u5bfc\u81f4\u516c\u5e73\u6027\u65b9\u9762\u6709\u4e89\u8bae\u7684\u6539\u8fdb(31)\u3002\u56e0\u6b64\uff0c\u7814\u7a76\u4eba\u5458\u5efa\u8bae\u8003\u8651\u66f4\u5e7f\u6cdb\u7684\u793e\u4f1a\u6280\u672f\u80cc\u666f(31)\uff0c\u867d\u7136\u8fd9\u53ef\u80fd\u7834\u574f\u7b97\u6cd5\u4e4b\u5916\u7684\u516c\u5e73\u6027\u3002\u5728\u6211\u4eec\u7684\u7814\u7a76\u8303\u56f4\u4e4b\u5916\uff0c\u8981\u8ba4\u8bc6\u5230\u516c\u5e73\u6027\u4e0d\u4ec5\u5728\u7b97\u6cd5\u7684\u5e94\u7528\u4e2d\u5f88\u91cd\u8981\uff0c\u800c\u4e14\u5728\u6574\u4e2a\u7814\u7a76\u8fc7\u7a0b\u4e2d\u4e5f\u5f88\u91cd\u8981\u3002\u4ece\u516c\u5e73\u7684\u7814\u7a76\u95ee\u9898\u3001\u516c\u5e73\u7684\u6570\u636e\u6536\u96c6\u3001\u516c\u5e73\u7684\u6570\u636e\u5206\u6790\u5230\u673a\u5668\u5b66\u4e60\u7b97\u6cd5\u7684\u90e8\u7f72\u540e\u8bc4\u4f30\u7684\u8fd9\u4e00\u6d41\u7a0b\uff0c\u5df2\u7ecf\u88ab\u5b9a\u4e49\u4e3a\u9a8c\u8bc1\u51b3\u7b56\u7b97\u6cd5\u7684\u201c\u516c\u5e73\u6027\u201d(8)\u3002\u6570\u636e\u4e0d\u5e73\u7b49\uff0c\u5373\u751f\u7269\u533b\u5b66\u4fdd\u5065\u7814\u7a76\u9886\u57df\u4e2d\u5c11\u6570\u7fa4\u4f53\u6210\u5458\u7684\u5c0f\u6216\u4f4e\u8d28\u91cf\u7684\u6837\u672c\u53ef\u4ee5\u88ab\u89e3\u51b3\uff1a\u8bad\u7ec3\u591a\u6570\u7fa4\u4f53\u6570\u636e\uff0c\u800c\u540e\u4f7f\u7528\uff08\u77e5\u8bc6\uff09\u8fc1\u79fb\u5b66\u4e60\u6765\u5fae\u8c03\u6a21\u578b\u5728\u5c11\u6570\u7fa4\u4f53\u6570\u636e\u4e2d\u7684\u8868\u73b0(11)\u3002\u7136\u800c\uff0c\u9664\u975e\u536b\u751f\u4fdd\u5065\u6570\u636e\u4e2d\u7684\u6570\u636e\u4e0d\u5e73\u7b49\u5f97\u5230\u7cfb\u7edf\u7684\u6539\u5584(11)\uff0c\u8fd9\u4ecd\u7136\u5b58\u5728\u7740\u4e00\u4e9b\u95ee\u9898\u3002<\/span><\/section>\n

<\/span><\/section>\n

\n
3.4.3 \u6cdb\u5316\u6027\u6216\u8005\u5916\u90e8\u6709\u6548\u6027<\/span><\/strong><\/section>\n

<\/span><\/section>\n<\/h4>\n
\u5916\u90e8\u6709\u6548\u7684\u6cdb\u5316\u80fd\u529b\u7684\u91cd\u8981\u6027\u4e0d\u4ec5\u9650\u4e8eML\u65b9\u6cd5\uff1a\u5728\u7814\u7a76\u8bbe\u8ba1\u548c\u6570\u636e\u6536\u96c6\u7684\u5404\u4e2a\u9636\u6bb5\uff0c\u56e0\u4e3a\u5e0c\u671b\u7814\u7a76\u7ed3\u679c\u7684\u6709\u6548\u6027\u8d85\u51fa\u53d1\u73b0\u5b83\u4eec\u7684\u6570\u636e\u96c6\uff0c\u90fd\u9700\u8981\u8003\u8651\u53ef\u80fd\u635f\u5bb3\u6cdb\u5316\u6027\u7684\u504f\u5dee\u3002\u5728 ML\u7684\u80cc\u666f\u4e0b\uff0c\u8003\u8651\u8fc7\u5ea6\u62df\u5408\u7684\u98ce\u9669\u662f\u91cd\u8981\u7684\uff0c\u5373\u901a\u8fc7\u672a\u4f7f\u7528\u6570\u636e\u96c6\u7684\u6a21\u578b\u6027\u80fd\u98ce\u9669\u63d0\u9ad8\u6570\u636e\u96c6\u5185\u6a21\u578b\u7684\u51c6\u786e\u6027\u3002<\/span><\/section>\n

<\/span><\/section>\n
\u4e00\u4e2a\u62db\u52df\u5dee\u5f02\u5982\u4f55\u5bfc\u81f4ML\u7b97\u6cd5\u6027\u80fd\u5dee\u5f02\u7684\u4f8b\u5b50\u662f\uff0832\uff09\u3002\u6b64\u5916\uff0c\u4e0e\u66f4\u4f20\u7edf\u7684\u5206\u6790\u7c7b\u4f3c\uff0c\u5efa\u8bae\u5bf9\u65b9\u6cd5\u8fdb\u884c\u4e09\u89d2\u6d4b\u91cf\u3002\u672c\u6587\u8fdb\u4e00\u6b65\u5f3a\u8c03\u5728\u7edf\u8ba1\u5206\u6790\u4e2d\u8bb0\u5f55\u6240\u6709\u51b3\u7b56\u7684\u91cd\u8981\u6027<\/span>\uff08\u4f8b\u5982\uff0c\u6307\u5b9a\u968f\u673a\u6570\u79cd\u5b50\uff09<\/span>\uff0c\u4ee5\u786e\u4fdd\u79d1\u5b66\u53d1\u73b0\u662f\u53ef\u590d\u5236\u7684\u3002<\/span><\/section>\n

<\/span><\/section>\n
\u6cdb\u5316\u6027\u7684\u4e00\u4e2a\u7279\u6b8a\u60c5\u51b5\u662f\u5173\u4e8e\u6d4b\u91cf\u7684\u5047\u8bbe\uff0c\u5373\u56e0\u7d20\u6216\u81ea\u53d8\u91cf\u4e0e\u611f\u5174\u8da3\u7684\u7ed3\u679c\u6216\u56e0\u53d8\u91cf\u662f\u9759\u6b62\u7684\uff0c\u5373\u4e0d\u968f\u65f6\u95f4\u53d8\u5316\uff0c\u4e5f\u4e0d\u53d6\u51b3\u4e8e\u6a21\u578b\u7684\u8f93\u5165\u3002\u7136\u800c\uff0c\u5728\u8bb8\u591a\u60c5\u51b5\u4e0b\uff0c\u9700\u8981\u610f\u8bc6\u5230\u6f5c\u5728\u7684\u6f02\u79fb\uff0c\u5373\u7531\u4e8e\u5916\u90e8\u56e0\u7d20\u800c\u5bfc\u81f4\u7684\u8f93\u5165\u4e0e\u8f93\u51fa\u95f4\u5173\u7cfb\u7684\u53d8\u5316\u3002\u8fd9\u4e0e\u4e0d\u65ad\u63a5\u6536\u65b0\u4f20\u5165\u6570\u636e\u8fdb\u884c\u5904\u7406\u7684ML\u5f15\u64ce\u76f8\u5173\uff0c\u8fd9\u4e9b\u6570\u636e\u53ef\u4ee5\u5229\u7528\u533b\u7597\u4fdd\u5065\u3001\u6559\u80b2\u6216\u6d88\u8d39\u8005\u6570\u636e\u8fdb\u884c\u8bbe\u7f6e<\/span>\uff08\u5c3d\u7ba1\u5728\u64b0\u5199\u672c\u6587\u65f6\uff0c\u5728\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u7814\u7a76\u4e2d\u5e76\u4e0d\u5e7f\u6cdb\u53ef\u7528\uff09<\/span>\u3002\u968f\u65f6\u95f4\u6f02\u79fb\u7684\u4e00\u4e2a\u4f8b\u5b50\u662f\uff0c\u5f53\u8d2b\u7a77\u7684\u793e\u4f1a\u51b3\u5b9a\u56e0\u7d20\u968f\u65f6\u95f4\u53d8\u5316\u65f6\uff0c\u8d2b\u56f0\u89e3\u91ca\u513f\u7ae5\u5065\u5eb7\u7684\u9884\u6d4b\u80fd\u529b\u964d\u4f4e\uff0c\u8bf7\u53c2\u9605\uff0833\uff09\u4e2d\u7684\u65e0\u969c\u788d\u4ecb\u7ecd\u548c\uff0834\uff09\u4e2d\u7684\u6700\u65b0\u53d1\u5c55\u3002\u867d\u7136\u8fc4\u4eca\u4e3a\u6b62\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u7684\u5927\u591a\u6570\u7814\u7a76\u90fd\u6d89\u53ca\u9759\u6001\u6570\u636e\u96c6\uff0c\u4f46\u968f\u7740\u4e0d\u65ad\u5b9e\u65f6\u5728\u7ebfML \u5f15\u64ce\u7684\u5f00\u53d1\uff0c\u5982\u8ddf\u8e2a\u5e94\u7528\u7a0b\u5e8f\u6216\u793e\u4ea4\u7f51\u7edc\u6570\u636e\uff0c\u8fd9\u4e00\u6982\u5ff5\u7684\u91cd\u8981\u6027\u5c06\u589e\u52a0\u3002<\/span><\/section>\n

<\/span><\/section>\n

\n
3.4.4 \u63d0\u5347\u4eba\u7c7b\u80fd\u529b\u548c\u6280\u80fd<\/span><\/strong><\/section>\n

<\/span><\/section>\n<\/h4>\n
\u6709\u8bb8\u591a ML \u53ef\u4ee5\u5728\u5de5\u4f5c\u6216\u65e5\u5e38\u751f\u6d3b\u4e2d\u652f\u6301\u4eba\u7c7b\u3002\u63a8\u8350\u7cfb\u7edf\u5e94\u7528\u4e8e\u591a\u4e2a\u9886\u57df\uff0c\u6839\u636e\u5148\u524d\u7684\u4fe1\u606f<\/span>\uff08\u4f8b\u5982\u76f8\u540c\u6216\u5176\u4ed6\u7528\u6237\u7684\u8bc4\u7ea7\uff09<\/span>\u5411\u7528\u6237\u63a8\u8350\u76f8\u5173\u9879\u76ee\u3002\u4ece\u5728\u65c5\u6e38\u3001\u5546\u4e1a\u548c\u8425\u9500\u4e2d\u5bf9\u63a8\u8350\u7cfb\u7edf\u7684\u4f7f\u7528\uff0c\u672c\u6587\u53ea\u662f\u4e3e\u4f8b\u8bf4\u660e\u63a8\u8350\u7cfb\u7edf\u5982\u4f55\u5e2e\u52a9\u5de5\u4f5c\u548c\u533b\u7597\u4fdd\u5065\uff1a\u63a8\u8350\u7cfb\u7edf\u53ef\u7528\u4e8e\u514b\u670d\u54a8\u8be2\u5e08\u4e0e\u81ea\u6740\u9884\u9632\u70ed\u7ebf\u53ef\u4ee5\u5e2e\u52a9\u4f5c\u5bb6\u5728\u987e\u95ee\u957f\u65f6\u95f4\u804a\u5929\u4e2d\u7684\u969c\u788d\u3002\u8be5\u65b9\u6cd5\u6709\u671b\u5e2e\u52a9\u60c5\u611f\u4e0a\u6709\u56f0\u96be\u8005\uff0c\u5373\u4fbf\u6b64\u65f6\u4eba\u7c7b\u4e13\u5bb6\u7684\u8868\u73b0\u4ecd\u7136\u8d85\u8fc7\u63a8\u8350\u7cfb\u7edf\uff0835\uff09\u3002\u6211\u4eec\u73b0\u5728\u8f6c\u5411\u7528\u4e8e\u63cf\u8ff0\u3001\u9884\u6d4b\u548c\u56e0\u679c\u63a8\u7406\u7684 ML \u65b9\u6cd5\u7684\u5206\u7c7b<\/span>\uff08\u5176\u4e2d\u4e00\u4e9b\u662f\u516c\u8ba4\u7684\u8d77\u6e90\u4e8e\u6807\u51c6\u7684\u7edf\u8ba1\u65b9\u6cd5\uff09<\/span>\uff0c\u5e76\u4ece\u63cf\u8ff0\u7528\u4e8e\u63cf\u8ff0\u7684ML\u65b9\u6cd5\u5f00\u59cb\u3002<\/span><\/section>\n

\n

<\/section>\n


<\/p>\n<\/h2>\n

\n
\n
\n
\n
\n

<\/section>\n

<\/section>\n
\n
<\/svg><\/section>\n<\/section>\n<\/section>\n
\n

4. \u63cf\u8ff0\u4e2d\u7684\u673a\u5668\u5b66\u4e60<\/strong><\/strong><\/p>\n<\/section>\n

\n

<\/section>\n

<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/h3>\n

<\/section>\n
\u4e00\u4e2a\u63cf\u8ff0\u6027\u7684\u7814\u7a76\u95ee\u9898\u65e8\u5728\u201c\u5bf9\u4e16\u754c\u7684\u67d0\u4e9b\u7279\u5f81\u63d0\u4f9b\u4e00\u4e2a\u5b9a\u91cf\u7684\u603b\u7ed3\u201d\uff086\uff09\u3002\u63cf\u8ff0\u662f\u6240\u6709\u5e94\u7528\u7814\u7a76\u7684\u57fa\u7840\uff0c\u56e0\u4e3a\u6211\u4eec\u9700\u8981\u63cf\u8ff0\u6765\u91cf\u5316\u6240\u7814\u7a76\u7684\u73b0\u8c61\uff0c\u4f8b\u5982\uff0c\u8bc4\u4f30<\/span>\uff08\u793e\u4f1a\uff09<\/span>\u7fa4\u4f53\u3001\u56fd\u5bb6\u6216\u5730\u7406\u5b9e\u4f53\u7684\u666e\u904d\u6027\u6216\u5206\u5e03\u53d8\u91cf\u4ee5\u53ca\u5176\u968f\u65f6\u95f4<\/span>\uff08\u961f\u5217\u4e4b\u95f4\uff09<\/span>\u7684\u53d8\u5316\u3002\u8fd9\u53ef\u4ee5\u901a\u8fc7\u51e0\u79cd\u7b97\u6cd5\u5b8c\u6210<\/span>\uff08\u5982\u88681\uff09<\/span>\u3002\u5728\u8f6c\u5411\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u7684\u7814\u7a76\u76ee\u6807\u4e0e\u63cf\u8ff0\u4e4b\u524d\uff0c\u6211\u4eec\u9996\u5148\u4ecb\u7ecd\u4e24\u7c7b\u76f8\u5173\u7684\u65b9\u6cd5\uff0c\u5373\u56e0\u5b50\u548c\u805a\u7c7b\u5206\u6790\uff0c\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u7684\u8bb8\u591a\u7814\u7a76\u4eba\u5458\u5df2\u7ecf\u610f\u8bc6\u5230\u4e86\u8fd9\u4e00\u70b9\u3002<\/span><\/section>\n

<\/span><\/section>\n
\u88681 \u7efc\u8ff0\u4e0e\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u4e2d\u6700\u76f8\u5173\u7684\u673a\u5668\u5b66\u4e60\u65b9\u6cd5\u975e\u6280\u672f\u63cf\u8ff0<\/span><\/section>\n\n\n\n\n\n\n\n\n\n\n\n\n\n
\n
\u7528\u4e8e\u63cf\u8ff0\u7684\u673a\u5668\u5b66\u4e60<\/span><\/strong><\/section>\n<\/td>\n<\/tr>\n
\n
\u805a\u7c7b<\/span><\/strong><\/section>\n<\/td>\n<\/tr>\n
\n
\u805a\u7c7b\u53ef\u8bc6\u522b\u6570\u636e\u4e2d\u7684\u76f8\u4f3c\u7ed3\u6784\uff0c\u5e76\u6839\u636e\u4e0e\u8ddd\u79bb\u3001\u5bc6\u5ea6\u76f8\u5173\u7684\u7279\u5b9a\u6807\u51c6\uff0c\u4ee5\u53ca\u805a\u96c6\u8f83\u5c0f\u7fa4\u4f53\u6216\u5212\u5206\u8f83\u5927\u7fa4\u4f53\u7684\u7b56\u7565\uff0c\u5c06\u4e0e\u4ee5\u524d\u672a\u77e5\u7684\u805a\u7c7b\u76f8\u4f3c\u7684\u5bf9\u8c61\u5206\u7ec4\u3002\u4f8b\u5b50\u5982\u4e0b:<\/span><\/section>\n
1\uff09K-means\uff1a\u6839\u636en\u7ef4\u7a7a\u95f4\u4e2d\u7684\u8ddd\u79bb\u5ea6\u91cf\u5bf9\u6570\u636e\u70b9\u8fdb\u884c\u805a\u7c7b<\/span><\/section>\n
2\uff09\u5c42\u6b21\u805a\u7c7b\uff1a\u4ece\u6bcf\u4e2a\u5bf9\u8c61\u5f62\u6210\u4e00\u4e2a\u5355\u72ec\u7684\u96c6\u7fa4\u5f00\u59cb\uff0c\u5411\u4e0a\u4f9d\u6b21\u5408\u5e76\u5f62\u6210\u96c6\u7fa4\u3002<\/span><\/section>\n
3\uff09\u57fa\u4e8e\u6a21\u578b\u7684\u805a\u7c7b\uff1a\u6700\u5e7f\u6cdb\u7684\u5e94\u7528\u7684\u662f\u9ad8\u65af\u6df7\u5408\u6a21\u578b<\/span>\uff08Gaussian mixture model, GMM\uff09<\/span>\uff0c\u5b83\u5141\u8bb8\u901a\u8fc7\u692d\u5706\u5f62\u805a\u7c7b\u3002<\/span><\/section>\n
4\uff09\u57fa\u4e8e\u5bc6\u5ea6\u7684\u805a\u7c7b\u65b9\u6cd5(DBSCAN)(131,132)\u5c06\u7a7a\u95f4\u90bb\u8fd1\u7684\u6570\u636e\u70b9\u5206\u4e3a\u4e00\u7ec4\uff0c\u540c\u65f6\u5c06\u4f4e\u5bc6\u5ea6\u533a\u57df\u4e2d\u7684\u6570\u636e\u70b9\u6807\u8bb0\u4e3a\u5f02\u5e38\u503c\u3002<\/span><\/section>\n
5\uff09\u4e13\u5bb6<\/span>\uff08\u5b66\u4e60\u8005\uff09<\/span>\u6df7\u5408\uff1a\u8be5\u65b9\u6cd5\u5c06\u8f93\u5165\u7a7a\u95f4\u5212\u5206\u4e3a\u540c\u8d28\u533a\u57df\uff0c\u6bcf\u4e2a\u533a\u57df\u7531\u4e0d\u540c\u7684\u4e13\u5bb6\u8d1f\u8d23\u3002\u8be5\u65b9\u6cd5\u5141\u8bb8\u4e0d\u540c\u7684\u96c6\u7fa4\uff0c\u4f8b\u5982\u60a3\u8005\u5728y\u548cx\u4e4b\u95f4\u5177\u6709\u4e0d\u540c\u7684 (\u975e) \u7ebf\u6027\u5173\u7cfb(133,134)\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
\u964d\u7ef4<\/span><\/strong><\/section>\n<\/td>\n<\/tr>\n
\n
1\uff09\u4e3b\u6210\u5206\u5206\u6790\uff1a\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u7684\u7814\u7a76\u4eba\u5458\u5df2\u7ecf\u77e5\u9053\uff0c\u8be5\u65b9\u6cd5\u901a\u8fc7\u7ebf\u6027 (\u6b63\u4ea4) \u6295\u5f71\u63d0\u4f9b\u6570\u636e\u7684\u4f4e\u7ef4\u8fd1\u4f3c\/\u7f16\u7801 (\u5373\u4f4e\u7ef4\u7279\u5f81\u662f\u539f\u59cb\u7279\u5f81\u7684\u7ebf\u6027\u7ec4\u5408)\u3002<\/span><\/section>\n
2\uff09\u6982\u7387\u4e3b\u6210\u5206\u5206\u6790 (PPCA) : \u7528\u4e8e\u4ee3\u8c22\u7ec4\u5b66\u4e2d\u53d1\u73b0\u7684\u9ad8\u7ef4\u6570\u636e\u7684\u4e3b\u6210\u5206\u5206\u6790\u7684\u6982\u7387\u6a21\u578b\u516c\u5f0f\u3002PPCA\u53ef\u4ee5\u4e3aPCA\u7684\u96f6\u566a\u58f0(135)\u9650\u5236\u4e0b\u63d0\u4f9b\u89e3\u51b3\u65b9\u6848\u3002<\/span><\/section>\n
3\uff09\u56e0\u5b50\u5206\u6790: \u6b63\u5982\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u7684\u7814\u7a76\u4eba\u5458\u6240\u77e5\uff0c\uff0c\u5982\u57fa\u4e8e\u8c03\u67e5\u95ee\u5377\u7684\u6570\u636e\u5206\u6790\u8fc7\u7a0b\u4e2d\uff0c\u56e0\u5b50\u5206\u6790\u53ef\u4ee5\u88ab\u89c6\u4e3a\u4e00\u4e2a\u5141\u8bb8\u7ef4\u7279\u5b9a\u566a\u58f0\u7684\u5e7f\u4e49PPCA\u3002\u8be5\u65b9\u6cd5\u901a\u8fc7\u5c11\u91cf\u7684\u6f5c\u5728\u56e0\u7d20\u6765\u89e3\u91ca\u7ef4\u5ea6\u4e4b\u95f4\u7684\u76f8\u5173\u6027\u3002<\/span><\/section>\n
4\uff09\u72ec\u7acb\u6210\u5206\u5206\u6790\uff1a\u56e0\u5b50\u5206\u6790\u7684\u63a8\u5e7f\uff0c\u5141\u8bb8\u6f5c\u5728\u56e0\u5b50\u7684\u5206\u5e03\u4e3a\u4efb\u610f\u975e\u9ad8\u65af\u5206\u5e03\u3002<\/span><\/section>\n
5\uff09\u975e\u7ebf\u6027\u964d\u7ef4\uff1a\u5305\u542b\u6838PCA\u3001\u9ad8\u65af\u8fc7\u7a0b\u9690\u53d8\u91cf\u6a21\u578b\u548ct-SNE\u3002<\/span><\/section>\n
6\uff09\u751f\u6210\u5f0f\u5bf9\u6297\u7f51\u7edc\uff1a\u57fa\u4e8e\u6570\u636e\u76f8\u4f3c\u6027\u805a\u7c7b\u7684\u4efb\u52a1\u5206\u4e3a\u4e24\u6b65\uff1a\u9996\u5148\uff0c\u751f\u6210\u771f\u5b9e\u6570\u636e\u76f8\u4f3c\u7684\u65b0\u6570\u636e\uff1b\u5176\u6b21\uff0c\u7c7b\u4f3c\u4e8e\u6709\u76d1\u7763\u5b66\u4e60\uff0c\u533a\u5206\u6570\u636e\u662f\u771f\u5b9e\u8fd8\u662f\u65b0\uff08\u5047\uff09\u6570\u636e\u3002\u4e00\u65e6\u7b97\u6cd5\u4e0d\u80fd\u533a\u5206\u771f\u5b9e\u6570\u636e\u548c\u65b0\u6570\u636e\uff0c\u4efb\u52a1\u5c31\u4f1a\u505c\u6b62(136)\u3002<\/span><\/section>\n
7\uff09\u53d8\u5206\u81ea\u7f16\u7801\u5668\uff1a\u4f7f\u7528\u795e\u7ecf\u7f51\u7edc\u8fdb\u884c\u964d\u7ef4\uff0c\u5305\u62ec\u7f16\u7801\u548c\u89e3\u7801\uff08\u5c06\u6570\u636e\u6620\u5c04\u5230\u4f4e\u7ef4\u6f5c\u5728\u7a7a\u95f4\uff0c\u53cd\u4e4b\u4ea6\u7136\uff09\u3002VAE\u4f7f\u7528\u6982\u7387\u516c\u5f0f\u548c\u53d8\u5206\u63a8\u7406\u6765\u5b66\u4e60\u6f5c\u5728\u53d8\u91cf\u7684\u5206\u5e03\uff0c\u8fd9\u6837\u53ef\u4ee5\u907f\u514d\u8fc7\u62df\u5408\uff0c\u5e76\u5728\u6f5c\u5728\u7a7a\u95f4\u4e0a\u589e\u52a0\u7406\u60f3\u7684\u5c5e\u6027(137)\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
\u5f02\u5e38\u68c0\u6d4b<\/span><\/strong><\/section>\n<\/td>\n<\/tr>\n
\n
\u8fd9\u662f\u4e00\u4e2a\u8bc6\u522b\u504f\u79bb\u6b63\u5e38\u201c\u884c\u4e3a\u201d\u6570\u636e\u70b9\u7684\u8fc7\u7a0b\uff0c\u5373\u5728\u6574\u4e2a\u6570\u636e\u70b9\u7684\u80cc\u666f\u4e2d\u8bc6\u522b\u4e0d\u540c\u7684\u6570\u636e\u70b9\u3002\u5f02\u5e38\u503c\u53ef\u80fd\u8868\u660e\u4e00\u4e2a\u4e8b\u4ef6\u7684\u8d8a\u8f68\u884c\u4e3a\uff08\u4f8b\u5982\uff0c\u6d88\u8d39\u6570\u636e\u4e2d\u7684\u6b3a\u9a97\u94f6\u884c\u8f6c\u8d26\u6570\u636e\u6216\u6d88\u8d39\u8005\u5bb6\u5ead\u7ec4\u6210\u7684\u53d8\u5316\uff09\u3002\u5f02\u5e38\u68c0\u6d4b\u53ef\u4ee5\u662f\u793e\u4ea4\u7f51\u7edc\u5206\u6790\u7684\u4e00\u79cd\u7279\u6b8a\u5f62\u5f0f\uff08\u89c1\u4e0b\u6587\uff09\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
\u53cc\u805a\u7c7b<\/span><\/strong><\/section>\n<\/td>\n<\/tr>\n
\n
\u4f5c\u4e3a\u4e00\u4e2a\u8f83\u65b0\u7684\u53d1\u5c55\uff0c\u4eba\u4eec\u4e5f\u53ef\u80fd\u4f1a\u5bf9\u540c\u65f6\u805a\u7c7b\u6837\u672c\uff08\u4e2a\u4f53\uff09\u548c\u7279\u5f81\uff08\u53d8\u91cf\uff09\u611f\u5174\u8da3\u3002\u6240\u8c13\u7684\u53cc\u805a\u7c7b\u65b9\u6cd5\u540c\u65f6\u5bf9\u6837\u672c\u548c\u7279\u5f81\u8fdb\u884c\u805a\u7c7b\uff1b\u6709\u5173\u7684\u6700\u65b0\u7efc\u8ff0\uff0c\u8bf7\u53c2\u89c1(138)\u3002\u53cc\u805a\u7c7b\u7528\u4e8e\u751f\u7269\u4fe1\u606f\u5b66\uff0c\u4f8b\u5982\u6839\u636e\u57fa\u56e0\u5b50\u96c6\u4e0a\u7684\u8868\u8fbe\u8c31\u5bf9\u60a3\u8005\u8fdb\u884c\u805a\u7c7b(139)\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
\u793e\u4ea4\u7f51\u7edc\u5206\u6790<\/span><\/strong><\/section>\n<\/td>\n<\/tr>\n
\n
SNA\u53ef\u7528\u4e8e\u8bc4\u4f30\u4e0d\u540c\u7684\u6570\u636e\u70b9\uff08\u8282\u70b9\uff1b\u5982\u7528\u6237\u3001\u9009\u6c11\u3001\u540c\u4e8b\u548c\u7ec4\u7ec7\uff09\u4e4b\u95f4\u7684\u8fde\u63a5\u6216\u5173\u7cfb\uff08\u8fb9\uff09\u3002SNA\u7528\u7ed3\u6784\u6216\u57fa\u4e8e\u5185\u5bb9\u7684\u5ea6\u91cf\u6765\u63cf\u8ff0\u7f51\u7edc\uff0c\u5982(53)\u6240\u8ff0\u3002\u7ed3\u6784\u6d4b\u91cf\u7684\u4f8b\u5b50\u662f\u4e2d\u5fc3\u6027\uff0c\u5176\u80fd\u8bc4\u4f30\u7f51\u7edc\u4e2d\u8282\u70b9\u7684\u76f8\u5173\u6027\u6216\u7ed3\u6784\u91cd\u8981\u6027\uff0c\u5e76\u53ef\u5c06\u4e2d\u5fc3\u6027\u6d4b\u91cf\u63a8\u5e7f\u5230\u4e00\u7ec4\u8282\u70b9\uff0c\u5982\u5ea6\u4e2d\u5fc3\u6027\uff08\u8fde\u63a5\u66f4\u591a\u7684\u8282\u70b9\u6392\u540d\u66f4\u9ad8\uff09\u3001\u7279\u5f81\u5411\u91cf\u4e2d\u5fc3\u6027\uff08\u8c03\u6574\u90bb\u5c45\u7684\u91cd\u8981\u6027\uff09\u6216\u7fa4\u4f53\u4e2d\u5fc3\u6027\u3002\u6b64\u5916\uff0c\u518d\u6b21\u4e0d\u540c\u7684\u8bc4\u4f30\u662f\u53ef\u80fd\u7684\uff0c\u5982\u7fa4\u4f53\u5ea6\u4e2d\u5fc3\u6027\u3002\u57fa\u4e8e\u5185\u5bb9\u7684\u5206\u6790\u53ef\u4ee5\u63d0\u53d6\u7528\u6237\u753b\u50cf\u6216\u4f1a\u8bdd\u4e3b\u9898\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

<\/span><\/section>\n
\u964d\u7ef4\u53ef\u4ee5\u964d\u4f4e\u6570\u636e\u96c6\u7684\u590d\u6742\u5ea6\uff0c\u4ee5\u4fbf\u540e\u7eed\u8fdb\u884c\u66f4\u6709\u6548\u7684\u5206\u6790\u3002\u56e0\u5b50\u5206\u6790\u5c06\u4e3a\u6240\u5305\u542b\u7684\u53d8\u91cf\u63d0\u4f9b\u56e0\u5b50\u548c\u56e0\u5b50\u8f7d\u8377<\/span>\uff08factor loading\uff09<\/span>\u3002\u5728\u540e\u7eed\u7684\u5206\u6790\u4e2d\uff0c\u56e0\u5b50\u901a\u5e38\u88ab\u7528\u4f5c\u5bcc\u96c6\u540e\u4fe1\u606f\u7684\u53d8\u91cf\u3002\u8bfb\u8005\u5e94\u8be5\u6ce8\u610f\u5230\uff0c\u57fa\u4e8e\u8d1d\u53f6\u65af\u5efa\u6a21\u7684\u7b97\u6cd5\u53ef\u7528\u4e8e\u56e0\u5b50\u5206\u6790\u548c\u672c\u6587\u6240\u4ecb\u7ecd\u7684\u5176\u4ed6\u65b9\u6cd5\uff0c\u4f46\u672c\u6587\u4e0d\u4f1a\u8be6\u7ec6\u4ecb\u7ecd\u3002\u9886\u57df\u77e5\u8bc6\u5bf9\u4e8e\u9884\u9009\u53ef\u89e3\u91ca\u548c\u6709\u610f\u4e49\u7684\u56e0\u7d20\u5206\u6790\u53d8\u91cf\u662f\u5fc5\u8981\u7684\uff0c\u56e0\u4e3a\u805a\u7c7b<\/span>\uff08\u548c\u56e0\u5b50\uff09<\/span>\u5206\u6790\u4e0d\u80fd\u4ece\u6982\u5ff5\u4e0a\u533a\u5206\u53d8\u91cf\uff0c\u4f8b\u5982\u6570\u636e\u6765\u81ea\u4eba\u7c7b\u6216\u5176\u4ed6\u7684\u60c5\u51b5\u3002\u6b64\u5916\uff0c\u4e0d\u5efa\u8bae\u5c06\u4e2a\u4f53\u53d8\u91cf\u4e0e\u9ad8\u9636\u53d8\u91cf\u6df7\u5408\uff0c\u4ee5\u786e\u4fdd\u540e\u7eed\u5206\u6790\u4e2d\u56e0\u5b50\u7684\u53ef\u89e3\u91ca\u6027\uff0c\u4f8b\u5982\u5728\u56e0\u5b50\u5206\u6790\u4e2d\u4e0e\u73af\u5883\u6216\u90bb\u8fd1\u5730\u533a\u76f8\u5173\u3002\u53e6\u4e00\u4e2a\u6709\u7528\u7684\u6570\u636e\u805a\u7c7b\u662f\u6839\u636e\u76f8\u4f3c\u6027\u5bf9\u6570\u636e\u8fdb\u884c\u5206\u7ec4\u3002\u4e0e\u56e0\u5b50\u5206\u6790\u4e0d\u540c\uff0c\u6570\u636e\u805a\u7c7b\u53ef\u4ee5\u540c\u65f6\u5e2e\u52a9\u5e94\u5bf9\u4e2a\u4f53\u6c34\u5e73\u548c\u80cc\u666f\u6c34\u5e73\u4e2d\u53d8\u91cf\u7684\u6d4b\u91cf\u3002\u4f8b\u5982\uff0c\u4e2a\u4f53\u5c42\u6b21\u7684BMI\u548c\u80cc\u666f\u5c42\u6b21\u7684\u7a7a\u6c14\u6c61\u67d3\uff0c\u4ee5\u8c03\u67e5\u5fc3\u8840\u7ba1\u6b7b\u4ea1\u7387\u7684\u4e0d\u540c\u98ce\u9669\u7ec4\u6216\u6982\u51b5<\/span>\uff08\u4f8b\u5982\u9ad8BMI\u548c\u9ad8\u7a7a\u6c14\u6c61\u67d3\uff09<\/span>\u3002\u6700\u540e\uff0c\u5982\u679c\u6837\u672c\u548c\u53d8\u91cf\u9700\u8981\u540c\u65f6\u5206\u7ec4\uff0c\u53cc\u805a\u7c7b\u53ef\u80fd\u4f1a\u6709\u6240\u5e2e\u52a9\u3002<\/span><\/section>\n

<\/span><\/section>\n
\u63cf\u8ff0\u6027\u7814\u7a76\u76ee\u6807\u53ef\u4ee5\u5728\u4f17\u591a\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u4e2d\u88ab\u53d1\u73b0\uff1a\u5305\u62ec(i)\u7b5b\u67e5\u548c\u8bc6\u522b\u9ad8\u5371\u4e2a\u4f53\u6216\u66f4\u9ad8\u6c34\u5e73\u7684\u6a21\u5f0f\u3001(ii)\u8bc6\u522b\u98ce\u9669\u6982\u51b5\u3001(iii)\u4e0d\u826f\u7ed3\u679c\u60a3\u75c5\u7387\u8fdb\u884c\u4f30\u8ba1\u548c\u9884\u6d4b\u4ee5\u53ca(iv)\u8bca\u65ad\u7684\u7814\u7a76\u76ee\u6807\u3002\u8fd9\u56db\u4e2a\u7814\u7a76\u76ee\u6807\u5c06\u5728\u672c\u8282\u7684\u5176\u4f59\u90e8\u5206\u4e2d\u63cf\u8ff0\u3002<\/span><\/section>\n

<\/span><\/section>\n

\n
4.1 \u7b5b\u9009\u548c\u8bc6\u522b\u9ad8\u5371\u4eba\u7fa4\u6216\u66f4\u9ad8\u5c42\u6b21\u7684\u6a21\u5f0f<\/span><\/strong><\/span><\/section>\n

<\/span><\/section>\n<\/h3>\n
\u53ef\u4ee5\u901a\u8fc7\u6570\u636e\u7684\u81ea\u52a8\u5316\u5904\u7406\u5bf9\u4e2a\u4eba\u8fdb\u884c\u9010\u4e00\u7684\u98ce\u9669\u56e0\u7d20\u7b5b\u9009\u3002\u4f8b\u5982\uff0c\u53ef\u4ee5\u901a\u8fc7\u5904\u7406\u7535\u5b50\u5065\u5eb7\u8bb0\u5f55\u6765\u786e\u5b9a\u4e0d\u826f\u5065\u5eb7\u7ed3\u679c\u98ce\u9669\u8f83\u9ad8\u7684\u4eba\u3002\u8fd9\u4e2a\u7814\u7a76\u76ee\u6807\u53ef\u4ee5\u7528\u5f02\u5e38\u68c0\u6d4b\u7684\u7b97\u6cd5\u6765\u5b8c\u6210\u3002\u901a\u8fc7\u805a\u7c7b\u6216\u56e0\u5b50\u5206\u6790\uff0c\u53ef\u4ee5\u5206\u6790\u4e2a\u4f53\u4e4b\u95f4\u7684\u6a21\u5f0f\uff0c\u5e76\u540c\u65f6\u5206\u6790\u4e00\u7ec4\u53d8\u91cf\u3002\u4e3a\u4e86\u786e\u5b9a\u4e0e\u8870\u8001\u76f8\u5173\u7684\u53d1\u75c5\u60c5\u51b5\uff0c\u4f7f\u7528\u4e0d\u540c\u7684\u805a\u7c7b\u7b97\u6cd5\u5bf9\u5305\u542b278\u79cd\u9ad8\u8d1f\u62c5<\/span>\uff08 high-burden\uff09<\/span>\u75be\u75c5\u4fe1\u606f\u7684\u7535\u5b50\u5065\u5eb7\u8bb0\u5f55\u8fdb\u884c\u5206\u6790\uff0c\u4ee5\u6839\u636e\u75be\u75c5\u53d1\u75c5\u65f6\u7684\u5e74\u9f84\u5bf9\u75be\u75c5\u8fdb\u884c\u5206\u7ec4(37)\u3002\u53e6\u5916\uff0c\u4eceML\u5f97\u5230\u7684\u590d\u6742\u4fe1\u606f\u53ef\u4ee5\u901a\u8fc7\u4f20\u7edf\u7684\u753b\u56fe\u8fdb\u884c\u53ef\u89c6\u5316\uff0c\u5982\u6bcf\u4e2a\u75be\u75c5\u7c07\u7684\u75be\u75c5\u66f2\u7ebf\u7684\u5206\u5e03(37)\u3002<\/span><\/section>\n

<\/span><\/section>\n
\u5bf9\u65f6\u53d8\u53d8\u91cf\u8f68\u8ff9\u7684\u7eb5\u5411\u5206\u6790\u53ef\u80fd\u6709\u52a9\u4e8e\u66f4\u597d\u5730\u4e86\u89e3\u4ee5\u524d\u786e\u5b9a\u7684\u5371\u9669\u56e0\u7d20\u7684\u77ed\u671f\u6216\u957f\u671f\u53d8\u5316\uff0c\u6700\u597d\u662f\u81f3\u5c11\u6709\u4e94\u4e2a\u968f\u8bbf\u6d4b\u91cf\u3002\u8bb8\u591a\u793e\u4f1a\u548c\u5065\u5eb7\u73b0\u8c61\u5728\u4e00\u751f\u4e2d\u90fd\u4f1a\u53d1\u751f\u53d8\u5316\uff0c\u5176\u8f68\u8ff9\u53ef\u4ee5\u88ab\u8ba4\u4e3a\u662f\u6b63\u5e38\u7684\u53d1\u5c55<\/span>\uff08\u4f8b\u5982\uff0c\u5728\u513f\u7ae5\u548c\u9752\u5c11\u5e74\u65f6\u671f\u8ba4\u77e5\u80fd\u529b\u7684\u63d0\u9ad8\uff0c\u5728\u665a\u5e74\u65f6\u671f\u8ba4\u77e5\u80fd\u529b\u7684\u4e0b\u964d\uff09<\/span>\u3002\u7136\u800c\uff0c\u5176\u4ed6\u53d8\u5316\u53ef\u4ee5\u9884\u6d4b\u75be\u75c5<\/span>\uff08\u4f8b\u5982\uff0c\u5f3a\u70c8\u7684\u4f53\u91cd\u589e\u52a0\uff0c\u5f3a\u70c8\u7684\u4f53\u91cd\u51cf\u8f7b\uff0c\u6216\u4e2d\u5e74\u8ba4\u77e5\u80fd\u529b\u4e0b\u964d\uff09<\/span>\uff0c\u56e0\u6b64\u8fd9\u4e0e\u5e72\u9884\u548c\u5b9e\u8df5\u76f8\u5173\u3002\u5728\u9700\u8981\u6301\u7eed\u76d1\u6d4b\u548c\u5e72\u9884\u7684\u60c5\u51b5\u4e0b\uff0c\u77ed\u671f\u91cd\u590d\u8bc4\u4f30\u53ef\u80fd\u662f\u76f8\u5173\u7684\uff0c\u4f8b\u5982\u5728\u624b\u672f\u80cc\u666f\u4e0b(38) \uff0c\u76d1\u6d4b\u56e0\u7d20\u7684\u8f68\u8ff9\u53ef\u80fd\u4f1a\u5e26\u6765\u5bf9\u60a3\u8005\u72b6\u6001\u7684\u989d\u5916\u7406\u89e3\u3002\u5728\u957f\u671f\u98ce\u9669\u9884\u6d4b\u76f8\u5173\u7684\u75be\u75c5\u4e2d\u3001\u6216\u8005\u5728\u524d\u671f\u9700\u8981\u533a\u5206\u56e0\u7d20\u662f\u5426\u4e3a\u5371\u9669\u56e0\u7d20\u4e2d\u3001\u6216\u5728\u65e9\u671f\u75c7\u72b6\u7684\u60c5\u51b5\u4e0b\uff0c\u957f\u671f\u91cd\u590d\u8bc4\u4f30\u53ef\u80fd\u662f\u6709\u8da3\u7684\u3002\u5728\u957f\u671f\u76d1\u6d4b\u4e2d\uff0c\u901a\u8fc7\u6291\u90c1\u75c7\u72b6\u6570\u91cf\u7684K-means\u7b97\u6cd5\u6765\u8bc6\u522b\u6bcf\u4e2a\u6d4b\u91cf\u573a\u5408\u7684\u6291\u90c1\u8f68\u8ff9\uff0c\u4ee5\u533a\u5206\u65e9\u53d1\u578b\u6291\u90c1\u75c7\u548c\u665a\u53d1\u578b\u6291\u90c1\u75c7\uff0c\u5e76\u6d4b\u8bd5\u5b83\u4eec\u4e0e\u4e0d\u826f\u5927\u8111\u7ed3\u679c\u7684\u5173\u8054(39)\u3002\u53e6\u4e00\u9879\u7814\u7a76\u901a\u8fc7\u6a21\u7cca\u805a\u7c7b\u9996\u53d1\u7cbe\u795e\u75c5\u60a3\u8005\u7684\u6570\u636e\u6765\u786e\u5b9a\u201c\u4f18\u826f\u9884\u540e\u201d\u3001\u201c\u7f13\u89e3\u8fc7\u7a0b\u201d\u3001\u201c\u4e34\u5e8a\u6076\u5316\u201d\u548c\u201c\u6162\u6027\u8fc7\u7a0b\u201d\u7684\u56db\u79cd\u8f68\u8ff9\uff0c\u5176\u4e2d\u6709\u4e0d\u540c\u7684\u6076\u5316\u548c\u7f13\u89e3\u8fc7\u7a0b\u7684\u5371\u9669\u56e0\u7d20\u3002\u7c7b\u4f3c\u7684\u5206\u6790\u53ef\u4ee5\u5e2e\u52a9\u5f3a\u8c03\u4e0d\u65ad\u53d8\u5316\u98ce\u9669\u56e0\u7d20\u7684\u91cd\u8981\u6027\u53ca\u5176\u5728\u751f\u547d\u8fc7\u7a0b\u4e2d\u7684\u76f8\u4e92\u4f5c\u7528\u3002<\/span><\/section>\n

<\/span><\/section>\n
\u6b7b\u4ea1\u4f5c\u4e3a\u4e00\u79cd\u5177\u6709\u7ade\u4e89\u6027\u7684\u98ce\u9669\u9700\u8981\u5728\u8001\u9f84\u5316\u76f8\u5173\u75be\u75c5\u7684\u8c03\u67e5\u4e2d\u52a0\u4ee5\u8003\u8651\uff0c\u4f8b\u5982\u968f\u673a\u5b58\u6d3b\u68ee\u6797\u3002\u8be5\u65b9\u6cd5\u540c\u6837\u53ef\u4ee5\u5bf9\u65f6\u53d8\u98ce\u9669\u56e0\u7d20\u8fdb\u884c\u5efa\u6a21(41)\u3002\u5982\u679c\u91cd\u70b9\u662f\u65f6\u53d8\u6cbb\u7597\u7684\u77ed\u671f\u540e\u679c\uff0c\u5373\u5e0c\u671b\u5f97\u5230\u56e0\u679c\u7ed3\u8bba\uff0c\u90a3\u4e48\u5728\u6f5c\u5728\u7ed3\u679c\u6846\u67b6\u4e2d\u4f7f\u7528\u5df2\u5efa\u7acb\u7684\u65b9\u6cd5<\/span>\uff08\u5982\u8fb9\u9645\u7ed3\u6784\u6a21\u578b\u6216 g \u516c\u5f0f\uff09<\/span>\u6548\u679c\u66f4\u597d(42)\u3002<\/span><\/section>\n

<\/span><\/section>\n

\n
4.2 \u8bc6\u522b\u98ce\u9669\u72b6\u51b5<\/span><\/strong><\/span><\/section>\n

<\/span><\/section>\n<\/h3>\n
\u4e3a\u4e86\u8bc6\u522b\u98ce\u9669\u7684\u5c5e\u6027\uff0c\u5373\u8bc6\u522b\u4e00\u7ec4\u4e2a\u4f53\u5bf9\u4e00\u7ec4\u53d8\u91cf\u7684\u7279\u5b9a\u503c\uff0c\u53ef\u4ee5\u57fa\u4e8e\u4ee5\u524d\u5df2\u7ecf\u8bc6\u522b\u51fa\u98ce\u9669\u6216\u4fdd\u62a4\u56e0\u5b50\u7684\u6587\u732e\uff0c\u91c7\u7528\u53d1\u73b0\u673a\u5668\u5b66\u4e60\u6765\u63cf\u8ff0\u5e76\u964d\u4f4e\u590d\u6742\u5ea6\u3002\u901a\u8fc7\u4f7f\u7528\u91cd\u75c7\u76d1\u62a4\u5ba4\u73af\u5883\u4e2d\u5e38\u89c1\u7684\u4e25\u91cd\u7a0b\u5ea6\u8bc4\u5206\uff0c\u60a3\u8005\u5065\u5eb7\u72b6\u51b5\u8f68\u8ff9\u53ef\u7528\u65f6\u95f4\u5e8f\u5217\u6570\u636e\u4e2d\u7684\u591a\u79cd\u964d\u7ef4<\/span>\uff08\u548c\u9884\u6d4b\uff09<\/span>\u6280\u672f\u8fdb\u884c\u5206\u7c7b\uff0c\u5305\u62ec\u57fa\u4e8e\u5bc6\u5ea6\u7684\u805a\u7c7b\u7b97\u6cd5<\/span>\uff08DBSCAN\uff09<\/span>(43)\u3002\u8fd9\u4e9b\u5065\u5eb7\u72b6\u6001\u8f68\u8ff9\u4e0e\u836f\u7269\u548c\u6cbb\u7597\u76f8\u5173\uff0c\u4f46\u5176\u7ed3\u679c\u6a21\u5f0f\u7684\u53ef\u89c6\u5316\u662f\u503c\u5f97\u80af\u5b9a\u7684 (43)\u3002<\/span><\/section>\n

<\/span><\/section>\n
\u867d\u7136\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u4e2d\u7684\u5927\u591a\u6570\u63cf\u8ff0\u6027\u7814\u7a76\u95ee\u9898\u90fd\u9700\u8981\u65e0\u76d1\u7763\u7684\u673a\u5668\u5b66\u4e60\uff0c\u4f46\u5bf9\u4e8e\u65e0\u76d1\u7763\u7684\u673a\u5668\u5b66\u4e60\u6765\u8bf4\uff0c\u63cf\u8ff0\u95ee\u9898\u4e2d\u4e0d\u5b58\u5728\u4e00\u5bf9\u4e00\u7684\u53cc\u5c04\u3002\u63a5\u4e0b\u6765\u5c06\u4ecb\u7ecd\u9700\u8981\u6709\u76d1\u7763\u673a\u5668\u5b66\u4e60\u65b9\u6cd5\u8fdb\u884c\u9884\u6d4b\u7684\u63cf\u8ff0\u6027\u7814\u7a76\u95ee\u9898\u3002<\/span><\/section>\n

<\/span><\/section>\n

\n
4.3 \u4f30\u7b97\u548c\u9884\u6d4b\u793e\u4f1a\u6216\u5065\u5eb7\u7ed3\u679c\u7684\u666e\u904d\u6027<\/span><\/strong><\/span><\/section>\n

<\/span><\/section>\n<\/h3>\n
\u4f30\u8ba1\u75be\u75c5\u6d41\u884c\u7387\u662f\u91cf\u5316\u536b\u751f\u8d1f\u62c5\u3001\u5e72\u9884\u9700\u6c42\u4ee5\u53ca\u89c4\u5212\u536b\u751f\u548c\u793e\u4f1a\u4fdd\u5065\u672a\u6765\u7684\u57fa\u7840\u3002\u4f30\u8ba1\u6240\u7814\u7a76\u73b0\u8c61\u7684\u53d1\u751f\u7387\u3001\u53d1\u75c5\u7387\u6216\u6d41\u884c\u7387\uff0c\u4f8b\u5982\u7279\u5b9a\u4eba\u7fa4\u4e2d\u7684\u5065\u5eb7\u7ed3\u679c\uff0c\u5982\u7cd6\u5c3f\u75c5\uff0c\u53ef\u4ee5\u88ab\u8ba4\u4e3a\u662f\u63cf\u8ff0\u6027\u7684\u3002\u4f30\u8ba1\u7684\u9884\u6d4b\u66f4\u503e\u5411\u4e8e\u9884\u6d4b\u6027\u7684\u3002\u4e3e\u4e00\u4e2a\u7eb3\u5165\u5065\u5eb7\u89c6\u89d2\u7684\u793e\u4f1a\u51b3\u5b9a\u56e0\u7d20\u4f8b\u5b50\u3002\u6839\u636e\u5e74\u9f84\u3001\u6027\u522b\u548c\u79cd\u65cf\/\u6c11\u65cf\u5bf9\u975e\u4f20\u67d3\u6027\u75be\u75c5<\/span>\uff08NCD\uff09<\/span>\u7684\u6d41\u884c\u7387\u8fdb\u884c\u4f30\u8ba1<\/span>\uff08\u53ef\u901a\u8fc7\u4e2a\u4f53\u5c42\u9762\u7684\u6570\u636e\u6216\u6c47\u603b\u6570\u636e\uff09<\/span>\uff0c\u800c\u540e\u901a\u8fc7ML\u4f30\u8ba1\u672a\u77e5\u5730\u533a\u7684\u6d41\u884c\u7387\u3002\u901a\u8fc7LASSO\u4f30\u8ba1\u4e866\u79cd\u975e\u4f20\u67d3\u6027\u75be\u75c5\u7684\u5b58\u5728\uff0c\u5e76\u4ee5\u6700\u5c0f\u7684\u4eba\u53e3\u7edf\u8ba1\u6570\u636e\u96c6\u9884\u6d4b\u4e86\u7f8e\u56fd5\u4e2a\u5dde(44)\u7684\u975e\u4f20\u67d3\u6027\u75be\u75c5\u6d41\u884c\u7387\u3002\u4e0d\u540c\u5730\uff0cWang\u548cRodriguez (45)\u4f7f\u7528\u5bf9\u6570\u7ebf\u6027\u6a21\u578b\u5bf9\u7a7a\u95f4\u968f\u673a\u6548\u5e94\u4f5c\u4e3a\u5e7f\u4e49\u6a21\u7ccaLASSO<\/span>\uff08generalized fussed LASSO\uff09<\/span>\u7684\u60e9\u7f5a\u51fd\u6570\uff0c\u4ee5\u786e\u5b9a\u7a7a\u95f4\u75be\u75c5\u7c07\uff0c\u5373\u53d1\u751f\u7387\u9ad8\u4e8e\u9884\u671f\u7684\u5730\u533a\u3002\u5728\u8fd9\u79cd\u60c5\u51b5\u4e0b\uff0c\u513f\u79d1\u764c\u75c7\u5728\u4f5b\u7f57\u91cc\u8fbe\u5dde\u6709\u8f83\u9ad8\u7684\u53d1\u75c5\u7387\u3002\u53e6\u4e00\u9879\u7814\u7a76\u4f7f\u7528\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\uff0c\u66f4\u5177\u4f53\u5730\u8bf4\u662f\u957f\u671f\u77ed\u671f\u8bb0\u5fc6\u548c\u95e8\u63a7\u5faa\u73af\u5355\u5143\u7f51\u7edc\u6765\u76d1\u6d4b\u75be\u75c5\u6d3b\u52a8\u3002\u5728\u8fd9\u79cd\u60c5\u51b5\u4e0b\uff0c\u5728\u7f8e\u56fd\u6d41\u611f\u75be\u75c5\u7206\u53d1\u6709\u591a\u79cd\u7a7a\u95f4\u5206\u8fa8\u7387(46)\u3002<\/span><\/section>\n

<\/span><\/section>\n

\n
4.4 \u8bca\u65ad<\/span><\/strong><\/span><\/section>\n

<\/span><\/section>\n<\/h3>\n
\u63cf\u8ff0\u6027\u7814\u7a76\u7684\u76ee\u6807\u662f\u786e\u5b9a\u4e00\u4e2a\u793e\u4f1a\u6216\u5065\u5eb7\u7ed3\u679c\u7684\u6d41\u884c\u7387\u3002\u4e0e\u5065\u5eb7\u72b6\u51b5\u6709\u5173\u7684\u8bca\u65ad\u901a\u5e38\u4e5f\u53ef\u4ee5\u5728\u6ca1\u6709\u4e34\u5e8a\u8bc4\u4f30\u7684\u60c5\u51b5\u4e0b\u8fdb\u884c\u5408\u7406\u63a8\u65ad\uff0c\u800c\u4e34\u5e8a\u8bc4\u4f30\u7684\u6307\u6807\u662f\u75be\u75c5\u6307\u6807<\/span>\uff08\u4f8b\u5982\uff0c\u7cd6\u5c3f\u75c5\u7684\u9ad8\u8840\u7cd6\uff09<\/span>\u3002\u6839\u636e\u73b0\u6709\u7684\u4fe1\u606f\uff0c\u63a8\u65ad\u51fa\u7684\u8bca\u65ad\u53ef\u4ee5\u88ab\u79f0\u4e3a\u201c\u53ef\u80fd\u201d\u3002\u5728\u63a5\u4e0b\u6765\u7684\u6587\u7ae0\u4e2d\uff0c\u6211\u4eec\u5c06\u4ecb\u7ecd\u4ee5\u4e0b\u76ee\u6807: (1)\u8bc4\u4f30\u53ef\u80fd\u7684\u8bca\u65ad\u3001(2)\u89e3\u51b3\u62a5\u544a\u4e0d\u8db3\u6216\u8bca\u65ad\u4e0d\u8db3\u3001\u4ee5\u53ca(3)\u901a\u8fc7\u7f51\u7edc\u8bc6\u522b\u548c\u4f20\u64ad\u793e\u4f1a\u6216\u6587\u5316\u5f71\u54cd\u3002<\/span><\/section>\n

<\/span><\/section>\n

\n
4.4.1 \u83b7\u53d6\u53ef\u80fd\u7684\u8bca\u65ad<\/span><\/strong><\/section>\n

<\/span><\/section>\n<\/h4>\n
\u7531\u4e8e\u8d44\u6e90\u7f3a\u4e4f\u6216\u6570\u636e\u9690\u79c1\u7684\u95ee\u9898\uff0c\u5728\u6570\u636e\u7a00\u7f3a\u7684\u73af\u5883\u4e2d\uff0c\u4e2a\u4f53<\/span>\uff08\u53d7\u8bbf\u8005\uff09<\/span>\u75be\u75c5\u7684\u63cf\u8ff0\u548c\u8bc6\u522b\u76ee\u6807\u53ef\u80fd\u662f\u76f8\u5173\u7684\u3002\u5982\u679c\u5206\u6790\u7684\u6570\u636e\u5305\u542b\u6709\u6807\u7b7e\u7684\u548c\u65e0\u6807\u7b7e\u7684\uff0c\u5219\u4efb\u52a1\u53ef\u4ee5\u901a\u8fc7\u6709\u76d1\u7763\u7684\u673a\u5668\u5b66\u4e60\u6216\u534a\u76d1\u7763\u6765\u5b9e\u73b0\u3002\u901a\u8fc7\u4f7f\u7528\u7535\u5b50\u5065\u5eb7\u8bb0\u5f55\u7684\u6570\u636e\uff0c\u5982\u795e\u7ecf\u7f51\u7edc\u7b49\u7b97\u6cd5\u53ef\u4ee5\u5728\u6ca1\u6709\u4eba\u7c7b\u4e34\u5e8a\u8bc4\u4f30\u7684\u60c5\u51b5\u4e0b\uff0c\u8bc6\u522b\u81f4\u75c5\u7684\u7279\u5f81\u7684\u5b58\u5728<\/span>\uff08\u6216\u8054\u5408\u5b58\u5728\u7684\u6761\u4ef6\uff09<\/span>\u3002\u5176\u4ed6\u6570\u636e\uff0c\u5982\u4f20\u611f\u5668\u6570\u636e\u548c\u8bed\u8a00\u6570\u636e\uff0c\u5219\u53ef\u4ee5\u901a\u8fc7\u672c\u6587\u672a\u6d89\u53ca\u7684\u65b9\u6cd5\u6765\u68c0\u6d4b\u6761\u4ef6\u6216\u75be\u75c5\uff0c\u5982\u81ea\u7136\u8bed\u8a00\u5904\u7406\u3002\u8bb8\u591a\u5176\u4ed6\u5e94\u7528\u5df2\u7ecf\u88ab\u7528\u4e8e\u68c0\u6d4b\u793e\u4ea4\u7f51\u7edc\u6570\u636e\u4e2d\u7684\u6291\u90c1\u75c7(47)\u3002<\/span><\/section>\n

<\/span><\/section>\n
\u5728\u7f3a\u4e4f\u57fa\u4e8e\u4e34\u5e8a\u8bc4\u4f30\u8bca\u65ad\u7684\u60c5\u51b5\u4e0b\uff0c\u901a\u8fc7ML\u5bf9\u6f5c\u5728\u60a3\u75c5\u7684\u4e2a\u4f53\u8fdb\u884c\u5206\u7c7b\u53ef\u6709\u52a9\u4e8e\u4f30\u8ba1\u75be\u75c5\u6d41\u884c\u7387\u548c\u76f8\u5173\u7684\u533b\u7597\u4fdd\u5065\u8d39\u7528\u3002\u867d\u7136\u786e\u5b9a\u6f5c\u5728\u60a3\u75c5\u4e2a\u4f53\u662f\u4e0d\u5f15\u4eba\u6ce8\u76ee\u7684\uff0c\u4f46\u6bd4\u5bf9\u8fd9\u4e9b\u4e2a\u4f53\u8fdb\u884c\u4e34\u5e8a\u8bc4\u4f30\u66f4\u5177\u6210\u672c\u6548\u76ca\u3002\u7ed3\u5408\u4e3b\u6210\u5206\u5206\u6790\u548c\u6570\u636e\u805a\u7c7b\u5206\u6790\uff0c\u7814\u7a76\u4eba\u5458\u5229\u7528\u4e16\u754c\u5404\u5730\u7684\u6570\u636e\u96c6\u5bf9\u75f4\u5446\u53ef\u80fd\u6027\u9ad8\u7684\u53c2\u4e0e\u8005\u8fdb\u884c\u4e86\u9274\u5b9a\u3002\u901a\u8fc7\u6587\u5316\u548c\u6559\u80b2\u516c\u5e73\u7684\u8ba4\u77e5\u6d4b\u8bd5\uff0c\u5c06ML\u5e94\u7528\u4e8e\u6765\u81ea\u5357\u5370\u5ea6\u7684\u6570\u636e(49)\uff0c10\/66\u7684\u75f4\u5446\u8bca\u65ad\u53ef\u4ee5\u5728\u6ca1\u6709\u4e34\u5e8a\u8bc4\u4f30\u7684\u60c5\u51b5\u4e0b\u901a\u8fc7\u8ba4\u77e5\u6d4b\u8bd5\u3002<\/span><\/section>\n

<\/span><\/section>\n

\n
4.4.2 \u89e3\u51b3\u62a5\u544a\u4e0d\u8db3\u6216\u8bca\u65ad\u4e0d\u8db3\u7684\u95ee\u9898<\/span><\/strong><\/section>\n

<\/span><\/section>\n<\/h4>\n
\u5bf9\u4e2a\u4f53\u7684\u6f5c\u5728\u8bca\u65ad\u8fdb\u884c\u8865\u5145\u5206\u7c7b\uff0c\u53ef\u4ee5\u89e3\u51b3\u75f4\u5446\u7b49\u75be\u75c5\u7684\u62a5\u544a\u4e0d\u8db3\u548c\u8bca\u65ad\u4e0d\u8db3\u7684\u95ee\u9898\u3002\u8fd9\u4e9b\u4e2a\u4f53\u6c34\u5e73\u7684\u53ef\u80fd\u8bca\u65ad\u53ef\u7528\u4e8e\u8c03\u67e5\u8fd9\u4e9b\u6761\u4ef6\u7684\u98ce\u9669\u548c\u4fdd\u62a4\u56e0\u7d20\u3002\u901a\u8fc7\u4f7f\u7528\u4f20\u7edf\u7b97\u6cd5\u548cML<\/span>\uff08\u63cf\u8ff0\u6027\u548c\u9884\u6d4b\u6027\uff09<\/span>\u7b97\u6cd5\u7684\u6df7\u5408\u7b97\u6cd5\uff0c\u8c03\u67e5\u53c2\u4e0e\u8005\u53ef\u80fd\u88ab\u786e\u8bca\u4e3a\u75f4\u5446\u3002\u8be5\u7b97\u6cd5\u7684\u76ee\u6807\u662f\u5728\u79cd\u65cf\/\u6c11\u65cf\u7fa4\u4f53\u4e2d\u63d0\u4f9b\u5177\u6709\u76f8\u4f3c\u654f\u611f\u6027\/\u7279\u5f02\u6027\u7684\u75f4\u5446\u5206\u7c7b\u7b97\u6cd5(50)\u3002\u8fd9\u9879\u7814\u7a76\u8bc1\u660e\u76f8\u6bd4\u4e8e\u4f20\u7edf\u7684\u4e13\u5bb6\u5efa\u6a21\uff0c\u5373\u57fa\u4e8e\u9886\u57df\u77e5\u8bc6\u7684\u7279\u5f81\u9009\u62e9\uff0c\u673a\u5668\u5b66\u4e60\u7684\u5b9e\u65bd\u66f4\u4e3a\u590d\u6742\u4e14\u5bf9\u961f\u5217\u548c\u7814\u7a76\u7a0b\u5e8f\u5dee\u5f02\u66f4\u654f\u611f; \u6b64\u5916\uff0c\u5728\u8fd9\u79cd\u60c5\u51b5\u4e0b\uff0c\u4f7f\u7528LASSO\u548c\u8d85\u7ea7\u5b66\u4e60\u5668\u7b49\u673a\u5668\u5b66\u4e60\u65b9\u6cd5\uff0c\u4e0e\u4e0d\u540c\u7684\u4e13\u5bb6\u6a21\u578b\u76f8\u6bd4\uff0c\u6ca1\u6709\u5e26\u6765\u6a21\u578b\u6027\u80fd\u7684\u589e\u52a0(50)\u3002\u56e0\u6b64\uff0c\u5bf9\u4e8e\u673a\u5668\u5b66\u4e60\u65b9\u6cd5\u7684\u4f7f\u7528\uff0c\u53ef\u80fd\u4e0d\u592a\u63a8\u8350\u4f7f\u7528\u65f6\u95f4\u8de8\u5ea6\u8f83\u957f\u7684\u4e0d\u52302000\u540d\u53c2\u4e0e\u8005\u7684\u6837\u672c\u3002<\/span><\/section>\n

<\/span><\/section>\n

\n
4.5 \u8bc6\u522b\u548c\u4f20\u64ad\u793e\u4f1a\u6216\u6587\u5316\u7684\u5f71\u54cd<\/span><\/strong><\/span><\/section>\n

<\/span><\/section>\n<\/h3>\n
\u4e00\u4e2a\u4e0e\u8bca\u65ad\u76f8\u5173\u7684\u7279\u522b\u7684\u7814\u7a76\u95ee\u9898\u53ef\u80fd\u662f\u5728\u7f51\u7edc\u4e2d\u8bc6\u522b\u548c\u4f20\u64ad\u793e\u4f1a\u6216\u6587\u5316\u5f71\u54cd\uff0c\u8fd9\u53ef\u4ee5\u7528\u793e\u4ea4\u7f51\u7edc\u5206\u6790<\/span>\uff08social network analysis, SNA\uff09<\/span>\u6765\u89e3\u51b3\u3002\u5728\u591a\u9879\u5b9e\u8bc1\u7814\u7a76\u4e2d\uff0c\u672c\u6587\u9009\u62e9\u51e0\u4e2a\u4f8b\u5b50\u4ee5\u8bf4\u660eSNA\u5982\u4f55\u6210\u4e3a\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u4e2d\u7ecf\u5178\u7edf\u8ba1\u5206\u6790\u7684\u6709\u7528\u5de5\u5177\uff0c\u5e76\u53ef\u7528\u4e8e\u5206\u6790\u793e\u793e\u4ea4\u7f51\u7edc\u548c\u4fe1\u4ef0\u4e2d\u653f\u6cbb\u6001\u5ea6\u4f20\u64ad\u7684\u51b3\u5b9a\u56e0\u7d20\uff0c\u4ee5\u53ca\u7ec4\u7ec7\u5185\u90e8\u548c\u7ec4\u7ec7\u4e4b\u95f4\u7684\u5408\u4f5c\u3002<\/span><\/section>\n

<\/span><\/section>\n
\u5728SNA\u7684\u6846\u67b6\u4e0b\uff0c\u5df2\u7ecf\u5f00\u53d1\u4e86\u7279\u5b9a\u7684ML\u65b9\u6cd5\uff0c\u5982\u6307\u6570\u65cf\u968f\u673a\u56fe\u6a21\u578b(51)\uff0c\u4f46\u662f\u6211\u4eec\u53d1\u73b0\u8bb8\u591a\u7528\u9884\u6d4b\u7684ML\u65b9\u6cd5\u8fdb\u884cSNA\u7684\u4f8b\u5b50\uff0c\u5982\u652f\u6301\u5411\u91cf\u673a\u6216\u6df1\u5ea6\u5b66\u4e60<\/span> (\u89c1\u88682)<\/span>\u3002SNA\u5df2\u7ecf\u88ab\u7528\u4e8e\u8861\u91cf\u653f\u6cbb\u503e\u5411\uff0c\u5e76\u53ef\u901a\u8fc7\u57fa\u4e8e\u4e0a\u4e0b\u6587\u7684\u5206\u6790\u6765\u8bc6\u522b\u7528\u6237\u6863\u6848\u548c\u5bf9\u8bdd\u4e3b\u9898\u3002Barber\u00e1 \u7b49\u4eba\u7684\u7814\u7a76\u662f\u5229\u7528\u8fd9\u4e24\u79cd\u65b9\u6cd5\u6765\u7406\u89e3\u5728\u793e\u4ea4\u5a92\u4f53\u4e0a<\/span>\uff08\u4f4e\u4e8e\u9884\u671f\u7684\uff09<\/span>\u653f\u6cbb\u503e\u5411\u3002SNA\u7684\u6982\u8ff0\u4ee5\u53ca\u8bc4\u4f30\u53ef\u4ee5\u5206\u4e3a\u7684\u56db\u4e2a\u4e0d\u540c\u7ef4\u5ea6\uff0c\u5373\u6a21\u5f0f\u548c\u77e5\u8bc6\u53d1\u73b0\u3001\u4fe1\u606f\u878d\u5408\u548c\u96c6\u6210\u3001\u53ef\u4f38\u7f29\u6027\u548c\u53ef\u89c6\u5316\uff0c\u8be6\u89c1(53)\u3002\u5e94\u7528\u4e8eSNA\u7684\u57fa\u4e8e\u4e0a\u4e0b\u6587\u7684ML\u65b9\u6cd5\u53ef\u4ee5\u8bc4\u4f30\u5176\u6027\u80fd\uff0c\u7c7b\u4f3c\u4e8e\u5728\u9884\u6d4b\u4e0a\u4e0b\u6587\u4e2d\u7684\u5e94\u7528\uff0c\u4f8b\u5982\u5173\u4e8e\u51c6\u786e\u6027\u6216\u7cbe\u786e\u6027(\u88684)\u3002\u5728\u63a5\u4e0b\u6765\u7684\u5c0f\u8282\u4e2d\uff0c\u6211\u4eec\u5c06\u63cf\u8ff0\u7528\u4e8e\u9884\u6d4b\u7684\u673a\u5668\u5b66\u4e60\u65b9\u6cd5\u3002<\/span><\/section>\n

<\/span><\/section>\n
\u88682 \u7efc\u8ff0\u4e0e\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u4e2d\u6700\u76f8\u5173\u7684\u673a\u5668\u5b66\u4e60\u65b9\u6cd5\u975e\u6280\u672f\u63cf\u8ff0<\/span><\/section>\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n
\n
\u7528\u4e8e\u9884\u6d4b\u7684\u673a\u5668\u5b66\u4e60<\/span><\/strong><\/section>\n<\/td>\n<\/tr>\n
\n
\u56de\u5f52<\/span><\/strong><\/section>\n<\/td>\n<\/tr>\n
\n
1\uff09\u7ebf\u6027\u56de\u5f52: \u901a\u8fc7\u5c06\u622a\u8ddd\u52a0\u5165\u9884\u6d4b\u56e0\u5b50\u53ca\u5176\u4f30\u8ba1\u7684\u52a0\u6743\u548c\u4e2d\u6765\u9884\u6d4b\u8fde\u7eed\u7684\u7ed3\u679c\uff0c\u6700\u5e38\u7528\u7684\u65b9\u6cd5\u662f\u4f7f\u9884\u6d4b\u8bef\u5dee\u7684\u5e73\u65b9\u4e00\u822c\u6700\u5c0f\u5e73\u65b9\u6cd5\u6700\u5c0f\u5316\u3002<\/span><\/section>\n
2\uff09\u903b\u8f91\u56de\u5f52: \u9884\u6d4b\u5206\u7c7b\u7ed3\u679c\u7684\u53ef\u80fd\u6027\uff0c\u4f7f\u7528\u903b\u8f91\u51fd\u6570\u4f5c\u4e3a\u5e7f\u4e49\u7ebf\u6027\u6a21\u5f0f\u4e2d\u7684\u8fde\u63a5\uff0c\u662f\u6700\u5e38\u7528\u7684\u7b26\u5408\u6700\u5927\u53ef\u80fd\u6027\u6216\u68af\u5ea6\u4e0b\u964d\u6cd5\uff0c\u5e76\u4e14\u53ef\u4ee5\u6269\u5c55\u5230\u53cd\u6620\u4e24\u4e2a\u4ee5\u4e0a\u7684\u7ed3\u679c\u7c7b\u522b\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
\u4eba\u5de5\u795e\u7ecf\u7f51\u7edc\u548c\u6df1\u5ea6\u5b66\u4e60<\/span><\/strong><\/section>\n<\/td>\n<\/tr>\n
\n
\u4eba\u5de5\u795e\u7ecf\u7f51\u7edc\u901a\u8fc7\u5927\u91cf\u7684\u6570\u636e\u70b9\uff08\u53d8\u91cf\u548c\u4e2a\u4f53\uff09\u6765\u641c\u7d22\u590d\u6742\u7684\u6570\u636e\u6a21\u5f0f\uff0c\u5efa\u7acb\u8fde\u63a5\u5355\u5143\u6216\u8282\u70b9\uff0c\u5e76\u6a21\u4eff\u7b80\u5316\u5927\u8111\u7ed3\u6784\u7684\u76f8\u4e92\u8fde\u63a5\u7684\u795e\u7ecf\u5143\u3002<\/span><\/section>\n
1\uff09\u5377\u79ef\u795e\u7ecf\u7f51\u7edc\uff1a\u5bf9\u8f83\u5c0f\u7684\u7f51\u683c\u6570\u636e\u5355\u5143\uff08\u5982\u6210\u50cf\u6216\u5e8f\u5217\u6570\u636e\uff09\u5e94\u7528\u8fc7\u6ee4\u5668\uff0c\u5e76\u5c06\u76f8\u5e94\u7684\u6fc0\u6d3b\u8f6c\u5316\u4e3a\u7279\u5f81\u6620\u5c04\uff0c\u4f46\u8ba1\u7b97\u4ee3\u4ef7\u9ad8\u4e14\u6570\u636e\u91cf\u5927\uff0c\u6027\u80fd\u4f18\u4e8e\u9700\u8981\u4eba\u5de5\u7279\u5f81\u5b9a\u4e49\u548c\u56fe\u50cf\u5206\u5272\u7684\u4f20\u7edf\u7684\u6210\u50cf\u5206\u6790\u65b9\u6cd5\u3002<\/span><\/section>\n
2\uff09\u591a\u5c42\u611f\u77e5\u673a\uff1a\u6700\u91cd\u8981\u7684\u4eba\u5de5\u795e\u7ecf\u7f51\u7edc\u7531\u4e00\u4e2a\u8f93\u5165\u5c42\u3001\u81f3\u5c11\u4e00\u4e2a\u9690\u542b\u5c42\u4ee5\u53ca\u4e00\u4e2a\u8f93\u51fa\u5c42\u7ec4\u6210\u3002\u6bcf\u4e00\u5c42\u7684\u795e\u7ecf\u5143\u5305\u542b\u524d\u4e00\u5c42\u795e\u7ecf\u5143\u7684\u52a0\u6743\u6c47\u603b\uff0c\u5e76\u7528\u6fc0\u6d3b\u51fd\u6570\u8ba1\u7b97\uff0c\u901a\u8fc7\u53cd\u5411\u4f20\u64ad\u63cf\u8ff0\u4e86\u5982\u4f55\u8c03\u6574\u6743\u91cd\u7684\u8fc7\u7a0b\uff0c\u4ee5\u5c3d\u53ef\u80fd\u5730\u51cf\u5c0f\u8f93\u51fa\u5c42\u7684\u8bef\u5dee\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
\u96c6\u6210\u65b9\u6cd5<\/span><\/strong><\/section>\n<\/td>\n<\/tr>\n
\n
1\uff09\u968f\u673a\u68ee\u6797\uff1a\u57fa\u4e8e\u8f93\u5165\u53d8\u91cf\u7684\u968f\u673a\u5b50\u96c6\uff0c\u5c06\u81ea\u5f15\u5bfc\u6982\u5ff5\u5e94\u7528\u5230\u4e2a\u4f53\u51b3\u7b56\u6811\u4e2d\uff0c\u6839\u636e\u4e2a\u4f53\u51b3\u7b56\u6811\u751f\u6210\u7684\u68ee\u6797\u6295\u7968\u5f97\u5230\u7ed3\u679c\u3002<\/span><\/section>\n
2\uff09XGBoost\uff1a\u68af\u5ea6\u63d0\u5347\u673a\u7684\u8f6f\u4ef6\u548c\u786c\u4ef6\u4f18\u5316\u5b9e\u73b0\uff0c\u4f7f\u5355\u4e2a\u51b3\u7b56\u6811\u7684\u987a\u5e8f\u7ec4\u5408\u7684\u9884\u6d4b\u8bef\u5dee\u6700\u5c0f\uff0c\u6027\u80fd\u4f18\u4e8e\u968f\u673a\u68ee\u6797\uff0c\u7279\u522b\u9002\u5408\u7ed3\u6784\u5316\u7684\u8868\u683c\u6570\u636e\u3002<\/span><\/section>\n
3\uff09\u8d1d\u53f6\u65af\u52a0\u6027\u56de\u5f52\u6811\uff1a\u5728\u8d1d\u53f6\u65af\u6846\u67b6\u4e2d\u5d4c\u5165\u5355\u4e2a\u51b3\u7b56\u6811\u7684\u987a\u5e8f\u96c6\u5408\uff0c\u5141\u8bb8\u5ea6\u91cf\u4f30\u8ba1\u503c\u7684\u4e0d\u786e\u5b9a\u6027\uff0c\u8bbe\u7f6e\u7edf\u8ba1\u7684\u9c81\u68d2\u9ed8\u8ba4\uff0c\u56e0\u6b64\u4e0d\u9700\u8981\u624b\u52a8\u6b63\u5219\u5316\u6216\u8c03\u6574\u8d85\u53c2\u6570\uff0c\u4e5f\u53ef\u4ee5\u7528\u4e8e\u56e0\u679c\u63a8\u7406\u3002<\/span><\/section>\n
4\uff09\u8d85\u7ea7\u5b66\u4e60\u5668\uff1a\u4f7f\u7528\u591a\u79cd\u63cf\u8ff0\u548c\u9884\u6d4b\u7684ML\u7b97\u6cd5\uff0c\u5e76\u901a\u8fc7\u4ea4\u53c9\u9a8c\u8bc1\u4f30\u8ba1\u5404\u81ea\u7684\u6027\u80fd\u3002\u8d85\u7ea7\u5b66\u4e60\u5668\u7684\u51c6\u786e\u6027\u63a5\u8fd1\u6700\u4f73\u9884\u6d4b\u7b97\u6cd5\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
\u652f\u6301\u5411\u91cf\u673a<\/span><\/strong><\/section>\n<\/td>\n<\/tr>\n
\n
\u652f\u6301\u5411\u91cf\u673a\uff08SVM\uff09\u4e00\u76f4\u662f\u5206\u7c7b\u6216\u56de\u5f52\u4e2d\u6700\u5e7f\u6cdb\u4f7f\u7528\u7684\u76d1\u7763\u5b66\u4e60\u65b9\u6cd5\u4e4b\u4e00\uff08140\uff09\u3002SVM\u901a\u8fc7\u5bfb\u627e\u6570\u636e\u70b9\uff08\u5411\u91cf\uff09\u4e4b\u95f4\u7684\u8d85\u5e73\u9762\u6765\u8bad\u7ec3\u6a21\u578b\uff0c\u5e76\u9009\u62e9\u6700\u80fd\u591f\u5c06\u4e00\u4e2a\u7c7b\u522b\u7684\u6570\u636e\u70b9\u4e0e\u53e6\u4e00\u4e2a\u7c7b\u522b\u7684\u6570\u636e\u70b9\u5206\u5f00\u7684\u8d85\u5e73\u9762\uff08\u5373\u5177\u6709\u6700\u5927\u7684\u4f59\u91cf\uff09\uff0c\u4e5f\u5c31\u662f\u8d85\u5e73\u9762\u4e0e\u6bcf\u4e2a\u7c7b\u522b\u7684\u6570\u636e\u70b9\u7684\u8ddd\u79bb\u662f\u6700\u5927\u7684\u3002\u5373\u4f7f\u5728\u8fd1\u5e74\u6765\u88ab\u68af\u5ea6\u4f18\u5316\u548c\u5176\u4ed6\u96c6\u6210\u7b97\u6cd5\uff08141\uff09\u8d85\u8d8a\uff0cSVM\u4ecd\u7136\u662f\u7a33\u5065\u7684\u9884\u6d4b\u65b9\u6cd5\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
\u6b63\u5219\u5316\u7b97\u6cd5<\/span><\/strong><\/section>\n<\/td>\n<\/tr>\n
\n
1) \u5cad\u56de\u5f52\uff1a\u5728\u4f3c\u7136\u4e4b\u548c\uff08\u6700\u5927\u4f3c\u7136\uff09\u6216\u5e73\u65b9\u4e4b\u548c\u4e2d\u52a0\u5165\u4e00\u4e2a\u57fa\u4e8e\u56de\u5f52\u7cfb\u6570\u7684\u6b63\u5219\u5316\u9879<\/span><\/section>\n
\u6b8b\u5dee\uff08\u666e\u901a\u6700\u5c0f\u4e8c\u4e58\u6cd5\uff09\uff0c\u53ef\u4ee5\u51cf\u5c0f\u5927\u578b\u6a21\u578b\u65b9\u5dee\u7684\u504f\u5dee\uff0c\u5373\u7f29\u5c0f\u4e86\u53c2\u6570\uff0c\u53ef\u5e94\u7528\u4e8e\u7ebf\u6027\u6216\u903b\u8f91\u56de\u5f52\u3002\u5f53\u5927\u591a\u6570\u53d8\u91cf\u662f\u6709\u4fe1\u606f\u65f6\uff0c\u7b97\u6cd5\u8868\u73b0\u826f\u597d\u3002<\/span><\/section>\n
2) LASSO\uff1a\u7c7b\u4f3c\u4e8e\u5cad\u56de\u5f52\uff0c\u5bf9\u53d8\u91cf\u9009\u62e9\u5f88\u6709\u7528\uff0c\u56e0\u4e3a\u5176\u53ef\u4ee5\u5c06\u975e\u4fe1\u606f\u53d8\u91cf\u7684\u4f30\u8ba1\u503c\u7f29\u51cf\u4e3a0\uff0c\u800c\u5cad\u56de\u5f52\u53ea\u80fd\u6e10\u8fdb\u5730\u5c06\u4f30\u8ba1\u503c\u54110\u6536\u7f29\u3002\u5f53\u8bb8\u591a\u53d8\u91cf\u4e3a\u975e\u4fe1\u606f\u53d8\u91cf\u65f6\uff0c\u7b97\u6cd5\u8868\u73b0\u826f\u597d\u3002<\/span><\/section>\n
3\uff09\u5f39\u6027\u7f51\uff1a\u5f15\u5165\u4e86\u57fa\u4e8eLASSO\u548c\u5cad\u56de\u5f52\u7684\u6b63\u5219\u5316\u3002\u5f53\u53d8\u91cf\u76f8\u4e92\u5173\u8054\u5e76\u4e14\u672a\u77e5\u53d8\u91cf\u4fe1\u606f\u91cf\u5927\u65f6\uff0c\u7b97\u6cd5\u6027\u80fd\u8868\u73b0\u826f\u597d\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
\u51b3\u7b56\u6811<\/span><\/strong><\/section>\n<\/td>\n<\/tr>\n
\n
1\uff09\u5206\u7c7b\u548c\u56de\u5f52\u6811\uff08CART\uff09\uff1a\u5355\u4e2a\u51b3\u7b56\u6811\u3002\u6570\u636e\u6839\u636e\u5206\u5272\u70b9\u6216\u5207\u5272\u70b9\u88ab\u53cd\u590d\u5206\u5272\u6210\u4e0d\u91cd\u53e0\u7684\u4ed3\u3002\u8fd9\u4e9b\u4ed3\u6307\u7684\u662f\u7528\u8d2a\u5a6a\u7b97\u6cd5\u786e\u5b9a\u7684\u9884\u6d4b\u5668\u503c\u3002\u9884\u6d4b\u662f\u57fa\u4e8e\u901a\u8fc7\u904d\u5386\u6811\u4e2d\u5404\u81ea\u8def\u5f84\u5230\u8fbe\u7684\u7ec8\u7aef\/\u53f6\u5b50\u8282\u70b9\u7684\u5e73\u5747\u503c\u6216\u591a\u6570\u7968\u3002\u968f\u673a\u68ee\u6797\u7b49\u57fa\u7840\u5143\u7d20\u3002<\/span><\/section>\n
2\uff09\u6761\u4ef6\u63a8\u7406\u6811\uff1a\u4e0eCART\u7c7b\u4f3c\u3002\u9012\u5f52\u5206\u533a\u7684\u53d8\u91cf\u9009\u62e9\u662f\u57fa\u4e8e\u663e\u8457\u6027\u68c0\u9a8c\uff0c\u800cCART\u662f\u6839\u636e\u4fe1\u606f\u6700\u5927\u5316\u63aa\u65bd\uff08\u5982\u57fa\u5c3c\u7cfb\u6570\uff09\u6765\u9009\u62e9\u5206\u5272\u53d8\u91cf\uff0c\u57fa\u5c3c\u7cfb\u6570\u662f\u6536\u5165\u4e0d\u5e73\u7b49\u7814\u7a76\u4e2d\u7ecf\u5e38\u4f7f\u7528\u7684\u7edf\u8ba1\u79bb\u6563\u6027\u7684\u8861\u91cf\u6807\u51c6\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

\n

<\/section>\n


<\/p>\n<\/h2>\n

\n
\n
\n
\n
\n

<\/section>\n

<\/section>\n
\n
<\/svg><\/section>\n<\/section>\n<\/section>\n
\n

5. \u9884\u6d4b\u4e2d\u7684\u673a\u5668\u5b66\u4e60<\/strong><\/strong><\/p>\n<\/section>\n

\n

<\/section>\n

<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/h3>\n

<\/section>\n
\u9884\u6d4b\u95ee\u9898\u5728\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u4e2d\u9ad8\u5ea6\u76f8\u5173\uff1a\u6211\u4eec\u53ef\u80fd\u5e0c\u671b\u67d0\u4e2a\u9884\u6d4b\uff0c\u5373\u793e\u4f1a\u6216\u5065\u5eb7\u7684\u7ed3\u679c\uff0c\u8981\u4e48\u5c3d\u53ef\u80fd\u51c6\u786e\uff0c\u8981\u4e48\u5c3d\u53ef\u80fd\u7b80\u660e\u3002\u7814\u7a76\u76ee\u6807\u53ef\u80fd\u662f\u89e3\u91ca\u7ed3\u679c\u4e2d\u7684\u6700\u5927\u65b9\u5dee\uff0c\u6216\u8005\u627e\u5230\u6700\u5c0f\u6216\u6700\u4f18\u7684\u9884\u6d4b\u56e0\u5b50\u96c6\uff0c\u4ee5\u6539\u5584\u5bf9\u9ad8\u5371\u4eba\u7fa4\u7684\u8bc6\u522b\u3002\u4e5f\u53ef\u80fd\u60f3\u8bc4\u4f30\u67d0\u4e2a\u5982\u6f5c\u5728\u98ce\u9669\u56e0\u7d20\u7684\u8f93\u5165\uff0c\u80fd\u591f\u591a\u597d\u5730\u9884\u6d4b\u7ed3\u679c\u3002\u9884\u540e\u7684\u6700\u7b80\u5355\u5f62\u5f0f\u662f\u9884\u6d4b\u3002\u53ef\u80fd\u6709\u786e\u5b9a\u7684\u7ec8\u70b9\uff0c\u6211\u4eec\u5e0c\u671b\u4f30\u8ba1\u8fbe\u5230\u5176\u4e2d\u4e00\u4e2a\u7ec8\u70b9\u7684\u6982\u7387\u3002\u4eceML\u7684\u89d2\u5ea6\u6765\u770b\uff0c\u8ba9\u8ba1\u7b97\u673a\/\u7b97\u6cd5\u4ee3\u66ff\u4eba\u7c7b\u6765\u5b9a\u4e49\u6a21\u578b\uff0cML\u53ef\u4ee5\u901a\u8fc7\u8003\u8651\u4e00\u5927\u6279\u53d8\u91cf\u6765\u6d4b\u8bd5\u4e00\u4e2a\u6216\u591a\u4e2a\u9884\u6d4b\u56e0\u5b50\u7684\u76f8\u5bf9\u91cd\u8981\u6027\uff0c\u5e76\u63d0\u4f9b\u56e0\u5b50\u91cd\u8981\u6027\u6216\u6392\u5e8f\u4fe1\u606f\u3002\u6b64\u5916\uff0c\u524d\u6587\u63d0\u5230\u7684\u201c\u7ef4\u5ea6\u8bc5\u5492\u201d\u4e5f\u540c\u6837\u9002\u7528\u3002\u5982\u679c\u60f3\u8981\u66f4\u4e3a\u6df1\u5165\u7684\u89e3\u91ca\uff0c\u53ef\u4ee5\u53c2\u8003\u9884\u6d4b\u5b66\u6559\u79d1\u4e66\uff0812\uff09\u3002<\/span><\/section>\n

<\/span><\/section>\n
\u9884\u6d4b\u7684\u7814\u7a76\u76ee\u6807\u9700\u8981\u5c06\u4e00\u4e9b\u7279\u5f81<\/span>\uff08\u8f93\u5165\uff09<\/span>\u5c3d\u53ef\u80fd\u51c6\u786e\u5730\u6620\u5c04\u5230\u5176\u4ed6\u5df2\u77e5\u7684\u7279\u5f81<\/span>\uff08\u8f93\u51fa\uff09<\/span>\u4e0a\uff086\uff09\u3002\u540c\u6837\uff0c\u5982\u5065\u5eb7\u7ed3\u679c\u7b49\u5df2\u77e5\u7ed3\u679c\uff0c\u88ab\u79f0\u4e3a\u6709\u6807\u7b7e\u6570\u636e\u3002\u56e0\u6b64\uff0c\u5728\u9884\u6d4b\u95ee\u9898\u4e2d\u4f7f\u7528\u6709\u76d1\u7763\u5b66\u4e60\u3002\u5982\u4e0a\u6240\u8ff0\uff0c\u6709\u76d1\u7763\u5b66\u4e60\u662f\u6307\u4ece\u6709\u6807\u7b7e\u7684\u6570\u636e\u4e2d\u5b66\u4e60\u4e00\u4e2a\u51fd\u6570\uff0c\u5c06\u201c\u8f93\u5165\u201d<\/span>\uff08\u9884\u6d4b\u56e0\u7d20\u548c\u81ea\u53d8\u91cf\uff09<\/span>\u6620\u5c04\u5230\u201c\u8f93\u51fa\u201d<\/span>\uff08\u7ed3\u679c\uff09<\/span>\u7684ML\u4efb\u52a1\u3002\u6570\u503c<\/span>\uff08\u8fde\u7eed\uff09<\/span>\u7ed3\u679c\u9700\u8981\u56de\u5f52\u6280\u672f\uff0c\u800c\u4e8c\u5206\u6216\u5206\u7c7b\u5219\u9700\u8981\u5206\u7c7b\u6280\u672f\u3002\u56fe2\u4ecb\u7ecd\u4e86ML\u65b9\u6cd5\u7684\u6982\u51b5\uff0c\u5728\u8003\u8651\u7406\u8bba\u548c\u7ecf\u9a8c\u7684\u57fa\u7840\u4e0a\uff0c\u5bf9\u8fd9\u4e9b\u65b9\u6cd5\u7684\u53ef\u89e3\u91ca\u6027\u4e0e\u590d\u6742\u6027\u4e4b\u95f4\u7684\u6743\u8861\u8fdb\u884c\u4e86\u6392\u5e8f\u3002\u5728\u4ecb\u7ecd\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u4e2d\u4e13\u95e8\u9488\u5bf9\u9884\u6d4b\u7684\u7814\u7a76\u95ee\u9898\u4e4b\u524d\uff0c\u672c\u8282\u5bf9\u901a\u5e38\u63d0\u5230\u7684\u7528\u4e8e\u9884\u6d4b\u7684\u5e38\u89c1ML\u65b9\u6cd5\u8fdb\u884c\u4e86\u6982\u8ff0\u3002<\/span><\/section>\n

<\/span><\/section>\n
<\/section>\n
\u673a\u5668\u5b66\u4e60\u65b9\u6cd5\u7528\u4e8e\u9884\u6d4b\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u7684\u975e\u6280\u672f\u63cf\u8ff0\uff0c\u6309\u53ef\u89e3\u91ca\u6027\/\u53ef\u89e3\u91ca\u6027\u4e0e\u590d\u6742\u6027\u8fdb\u884c\u6392\u5e8f\u3002<\/span><\/section>\n

<\/span><\/section>\n

\n
5.1 \u5e38\u89c1\u7684\u9884\u6d4b\u6027\u673a\u5668\u5b66\u4e60\u65b9\u6cd5<\/span><\/strong><\/span><\/section>\n

<\/span><\/section>\n<\/h3>\n
\u4f5c\u4e3a\u5c11\u6570\u6d41\u884c\u7684\u6216\u8868\u73b0\u826f\u597d\u7684ML\u65b9\u6cd5\uff0c\u89e3\u51b3\u9884\u6d4b\u95ee\u9898\u7684\u5178\u578bML\u65b9\u6cd5\u6709\uff1a\uff08i\uff09\u60e9\u7f5a\u6027\u65b9\u6cd5\u3001\uff08ii\uff09\u96c6\u5408\u5b66\u4e60\u4ee5\u53ca\uff08iii\uff09\u795e\u7ecf\u7f51\u7edc\uff0c\u4e0b\u9762\u5c06\u5bf9\u8fd9\u4e9b\u65b9\u6cd5\u8fdb\u884c\u66f4\u8be6\u7ec6\u7684\u6280\u672f\u63cf\u8ff0\uff0c\u66f4\u591a\u7684\u65b9\u6cd5\u5982\u88682\u6240\u793a\u3002<\/span><\/section>\n

<\/span><\/section>\n

\n
5.1.1 \u60e9\u7f5a\u6027\u7684\u65b9\u6cd5<\/span><\/strong><\/section>\n

<\/span><\/section>\n<\/h4>\n
\u7f29\u51cf\u6216\u60e9\u7f5a\u6027\u56de\u5f52\u65b9\u6cd5\u4e5f\u88ab\u7528\u4e8e\u4f20\u7edf\u7684\u63a8\u7406\u7edf\u8ba1\u4e2d\u3002\u60e9\u7f5a\u6027\u56de\u5f52\u5728\u5927\u578b\u591a\u53d8\u91cf\u6570\u636e\u96c6\u4e2d\u8868\u73b0\u5f97\u6bd4\u6807\u51c6\u7ebf\u6027\u6a21\u578b\u66f4\u597d\uff0c\u662f\u56e0\u4e3a\u8fd9\u4e9b\u6570\u636e\u96c6\u7684\u53d8\u91cf\u6bd4\u4e2a\u4f53\u66f4\u591a\u3002\u60e9\u7f5a\u6027\u56de\u5f52\u5c06\u5728\u65b9\u7a0b\u4e2d\u6dfb\u52a0\u4e00\u4e2a\u7ea6\u675f\u6761\u4ef6\uff0c\u4ee5\u60e9\u7f5a\u6a21\u578b\u4e2d\u53d8\u91cf\u8fc7\u591a\u7684\u7ebf\u6027\u56de\u5f52\u6a21\u578b\uff0c\u4e5f\u79f0\u4e3a\u201c\u6536\u7f29\u201d\u6216\u201c\u6b63\u5219\u5316\u201d\u3002\u8fd9\u5c06\u4f7f\u7cfb\u6570\u503c\u5411\u96f6\u7f29\u51cf\uff0c\u8fdb\u800c\u5bfc\u81f4\u8d21\u732e\u8f83\u5c11\u7684\u53d8\u91cf\u7684\u7cfb\u6570\u5c06\u63a5\u8fd1\u96f6\u6216\u7b49\u4e8e\u96f6\uff084\uff09\u3002\u6700\u5e7f\u6cdb\u4f7f\u7528\u7684\u60e9\u7f5a\u65b9\u6cd5\uff0c\u5bf9\u5e94\u4e8e\u4e0d\u540c\u7684\u6536\u7f29\u60e9\u7f5a\uff0c\u5305\u62ecLASSO\u3001\u5cad\u56de\u5f52\u548c\u5f39\u6027\u7f51\u3002\u7136\u800c\uff0c\u867d\u7136LASSO\u5c06\u4e00\u4e9b\u7cfb\u6570\u8bbe\u4e3a\u96f6\uff0c\u4ece\u800c\u540c\u65f6\u8fdb\u884c\u53d8\u91cf\u9009\u62e9\u548c\u4f30\u8ba1\uff0c\u4f46\u5176\u5b58\u5728\u7740\u4f17\u6240\u5468\u77e5\u7684\u4f30\u8ba1\u504f\u5dee\u3002\u4e3a\u4e86\u514b\u670d\u8fd9\u4e2a\u95ee\u9898\uff0c\u4eba\u4eec\u8bbe\u8ba1\u4e86\u5305\u62ec\u5e73\u6ed1\u526a\u5207\u7684\u7edd\u5bf9\u504f\u5dee\uff0854\uff09\u7684\u5176\u4ed6\u60e9\u7f5a\u63aa\u65bd\u3002\u6b64\u5916\uff0c\u8fd8\u63d0\u51fa\u4e86\u8003\u8651\u6570\u636e\u7ed3\u6784\u7684\u5404\u79cd\u6269\u5c55\uff0c\u5982\u7528\u4e8e\u65f6\u95f4\u7ed3\u6784\u53d8\u91cf<\/span>\uff08\u6216\u66f4\u666e\u904d\u7684\u56fe\u7ed3\u6784\u53d8\u91cf\uff09<\/span>\u7684fussed LASSO\u548c\u7528\u4e8e\u7ec4\u7ed3\u6784\u53d8\u91cf\u7684group LASSO\u3002<\/span><\/section>\n

<\/span><\/section>\n

\n
5.1.2 \u96c6\u6210\u5b66\u4e60<\/span><\/strong><\/section>\n

<\/span><\/section>\n<\/h4>\n
\u4e0b\u9762\u5c06\u4ecb\u7ecd\u5e38\u89c1\u7684\u96c6\u6210\u5b66\u4e60\u7b97\u6cd5\uff0c\u5305\u542b\u968f\u673a\u68ee\u6797\u3001XGBoost\u548c\u8d85\u7ea7\u5b66\u4e60\u5668\u3002BART\u4e5f\u662f\u4e00\u79cd\u96c6\u6210\u5b66\u4e60\u7b97\u6cd5\uff0c\u4f46\u5176\u76f8\u5173\u8f6f\u4ef6\u5305\u63d0\u4f9b\u4e86\u8bb8\u591a\u57fa\u4e8e\u9886\u57df\u77e5\u8bc6\u7684\u56e0\u679c\u5206\u6790\u8bbe\u7f6e\uff0855\uff0c56\uff09\uff0c\u56e0\u6b64\u5e38\u88ab\u7528\u4e8e\u7814\u7a76\u56e0\u679c\u95ee\u9898\uff0c\u6545\u5c06\u5728\u7b2c\u516d\u8282\u8fdb\u884c\u4ecb\u7ecd\u3002\u9996\u5148\uff0c\u5982\u968f\u673a\u68ee\u6797\uff0857\uff09\u7b49\u96c6\u6210\u65b9\u6cd5\u4ee5\u5176\u53ef\u89e3\u91ca\u6027\u5df2\u88ab\u7528\u4e8e\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u7684\u8bb8\u591a\u7814\u7a76\u4e2d\uff0c\u5982\uff0841\uff0c58\uff09\u3002<\/span><\/section>\n

<\/span><\/section>\n
\u5176\u6b21\uff0c\u7531\u4e8e\u5176\u9884\u6d4b\u7cbe\u5ea6\u9ad8\u548c\u5bf9\u8bb8\u591a\u9884\u6d4b\u95ee\u9898\u7684\u5f3a\u5927\u9002\u7528\u6027\uff0c\u672c\u6587\u5411\u8bfb\u8005\u4ecb\u7ecd\u968f\u673a\u68af\u5ea6\u63d0\u5347\uff0859\uff09\u3002Python\u548cR\u4e2d\u5df2\u7ecf\u5b9e\u73b0\u4e86\u68af\u5ea6\u63d0\u5347\u7b97\u6cd5\u7cfb\u5217\u7684\u5373\u7528\u578b\u8f6f\u4ef6\u5305\uff0c\u5982\u63a8\u8350\u4e2d\u7684\u201cXGBoost\u201d\u8f6f\u4ef6\u5305\u3002\u901a\u8fc7\u63d0\u5347\u7b97\u6cd5\uff0c\u5178\u578b\u7684\u8f93\u5165\u53d8\u91cf<\/span>\uff08\u9884\u6d4b\u56e0\u5b50\uff09<\/span>\u7684\u5171\u7ebf\u6027\u95ee\u9898\u5c31\u4e0d\u4f1a\u51fa\u73b0\u3002\u8fd9\u610f\u5473\u7740 \u4e0d\u540c\u7684\u5de5\u7a0b<\/span>\uff08\u9884\u5904\u7406\uff09<\/span>\u53d8\u91cf\u53ef\u4ee5\u540c\u65f6\u8f93\u5165\uff0c\u4ee5\u5bfb\u627e\u6700\u5177\u9884\u6d4b\u6027\u7684\u7279\u5f81\u3002\u7814\u7a76\u4eba\u5458\u9700\u8981\u6ce8\u610f\u7684\u662f\uff0c\u8be5\u7b97\u6cd5\u4f1a\u5728\u6240\u6709\u53d8\u91cf\u4e0a\u8be6\u5c3d\u5730\u641c\u7d22\u4ee5\u5bfb\u627e\u5206\u5272\u70b9\uff0c\u5e76\u4e14\u5728\u5212\u5206\u6837\u672c\u65f6\u6709\u4e9b\u53d8\u91cf\u53ef\u80fd\u6bd4\u5176\u4ed6\u53d8\u91cf\u66f4\u6709\u53c2\u8003\u4ef7\u503c\u3002\u56e0\u6b64\uff0c\u8be5\u7b97\u6cd5\u504f\u5411\u4e8e\u9009\u62e9\u6570\u503c\u7684<\/span>\uff08\u8fde\u7eed\u7684\uff09<\/span>\u3001\u591a\u7c7b\u522b\u7684\u53d8\u91cf\u6216\u6709\u7f3a\u5931\u6570\u636e\u7684\u53d8\u91cf\u800c\u4e0d\u662f\u4e8c\u5206\u6cd5\u53d8\u91cf\u3002\u8be5\u7b97\u6cd5\u53ef\u4ee5\u8fdb\u884c\u65e0\u504f\u7684\u53d8\u91cf\u9009\u62e9\uff0860\uff09\u3002\u6b64\u5916\uff0c\u90a3\u4e9b\u6700\u80fd\u89e3\u91ca\u56e0\u53d8\u91cf\u7684\u53d8\u91cf\u5c06\u88ab\u9009\u4e3a\u5206\u5272\u70b9\uff0c\u4f46\u4ece\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u7684\u76f8\u5173\u7406\u8bba\u89d2\u5ea6\u6765\u770b\uff0c\u8fd9\u4e0d\u4e00\u5b9a\u662f\u6700\u6709\u610f\u4e49\u7684\u7406\u8bba\u53d8\u91cf\u3002<\/span><\/section>\n

<\/span><\/section>\n
\u6700\u540e\uff0c\u4e00\u4e9b\u96c6\u6210\u5b66\u4e60\u7b97\u6cd5\u6d4b\u8bd5\u4e86\u8de8\u591a\u4e2aML\u7b97\u6cd5\u5f00\u53d1\u7684\u6a21\u578b\u6027\u80fd\u7684\u53ef\u80fd\u6027\u3002\u7b97\u6cd5\u5bb6\u65cf\u4e2d\u7684\u4e00\u4e2a\u4f8b\u5b50\u662f\u8d85\u7ea7\u5b66\u4e60\u5668\uff0c\u5176\u5df2\u88ab\u5e94\u7528\u4e8e\u6d41\u884c\u75c5\u5b66\u7814\u7a76\u95ee\u9898\uff0861\uff09\u3002\u8d85\u7ea7\u5b66\u4e60\u5668\u4f7f\u7528\u4ea4\u53c9\u9a8c\u8bc1\u6765\u4f30\u8ba1\u51e0\u4e2a\u63cf\u8ff0\u6027\u548c\u9884\u6d4b\u6027ML\u6a21\u578b\u7684\u6027\u80fd\uff0c\u6216\u8005\u5728\u540c\u4e00\u6a21\u578b\u4e2d\u4f7f\u7528\u4e0d\u540c\u7684\u8bbe\u7f6e\u53c2\u6570\uff0c\u5e76\u4e14\u5728\u6a21\u578b\u62df\u5408\u8fc7\u7a0b\u4e2d\u4e0e\u4f7f\u7528\u7684\u6700\u4f73\u9884\u6d4b\u7b97\u6cd5\u4e00\u6837\u5730\u903c\u8fd1\u51c6\u786e\u503c\u3002\u4e00\u4e2a\u8fd9\u65b9\u9762\u7684\u5e94\u7528\u662f\u7f3a\u5931\u6570\u636e\u7684\u8865\u5168\uff0862\uff09\u3002<\/span><\/section>\n

<\/span><\/section>\n

\n
5.1.3 \u4eba\u5de5\u795e\u7ecf\u7f51\u7edc<\/span><\/strong><\/section>\n

<\/span><\/section>\n<\/h4>\n
\u4eba\u5de5\u795e\u7ecf\u7f51\u7edc\u901a\u5e38\u5229\u7528\u5927\u91cf\u7684\u6570\u636e\u6765\u641c\u7d22\u590d\u6742\u7684\u6a21\u5f0f\u5e76\u5efa\u7acb\u8fde\u63a5\u5355\u5143\u6216\u8282\u70b9\u3002\u5177\u6709\u4e00\u4e2a\u4ee5\u4e0a\u9690\u85cf\u5c42\u7684\u4eba\u5de5\u795e\u7ecf\u7f51\u7edc\u4e5f\u88ab\u79f0\u4e3a\u6df1\u5ea6\u5b66\u4e60\u3002<\/span><\/section>\n

<\/span><\/section>\n
\u4eba\u5de5\u795e\u7ecf\u7f51\u7edc\u8be6\u7ec6\u7684\u63cf\u8ff0\uff1a\u901a\u8fc7\u5c06\u4e00\u4e9b\u7279\u5f81<\/span>\uff08\u9884\u6d4b\u56e0\u5b50\uff09<\/span>\u6620\u5c04\u5230\u5176\u4ed6\u7279\u5f81<\/span>\uff08\u7ed3\u679c\uff09<\/span>\u6765\u89e3\u51b3\u9884\u6d4b\u95ee\u9898\u7684\u4e00\u79cd\u66f4\u590d\u6742\u7684\u65b9\u6cd5\u53ef\u4ee5\u7528\u795e\u7ecf\u7f51\u7edc\u6765\u5b8c\u6210\uff0c\u5176\u5177\u6709\u5b9a\u4e49\u7684\u8f93\u5165<\/span>\uff08\u9884\u6d4b\u56e0\u7d20\uff09<\/span>\u548c \u8f93\u51fa<\/span>\uff08\u7ed3\u679c\uff09<\/span>\uff0c\u5e76\u5b66\u4e60\u53d8\u91cf\u95f4\u7684\u76f8\u4e92\u5173\u7cfb\u3002\u795e\u7ecf\u7f51\u7edc\u662f\u53d7\u5927\u8111\u4e2d\u7684\u8ba1\u7b97\u4e3a\u9a71\u52a8\uff0c\u5e76\u4f7f\u590d\u6742\u4efb\u52a1\u7684\u6210\u529f\u8bc6\u522b\u548c\u5206\u7c7b\u6210\u4e3a\u53ef\u80fd\uff0863\uff09\u3002\u5982\u679c\u91c7\u7528\u6700\u5e38\u89c1\u7684\u903b\u8f91\u6fc0\u6d3b\u51fd\u6570\uff0864\uff09\uff0c\u795e\u7ecf\u7f51\u7edc\u4e0e\u6ca1\u6709\u9690\u542b\u5c42\u7684\u4f20\u7edf\u903b\u8f91\u56de\u5f52\u662f\u4e00\u6837\u7684\u3002\u795e\u7ecf\u7f51\u7edc\u548c\u903b\u8f91\u56de\u5f52\u90fd\u6709\u4e00\u4e2a\u51fd\u6570\u5f62\u5f0f\uff0c\u5e76\u4e14\u53c2\u6570\u662f\u7531\u6700\u5927\u4f3c\u7136\u4f30\u8ba1\u786e\u5b9a\u7684\u3002\u7136\u800c\uff0c\u795e\u7ecf\u7f51\u7edc\u5141\u8bb8\u6a21\u578b\u653e\u5bbd\u5173\u4e8e\u8f93\u5165\u53d8\u91cf\u7ebf\u6027\u548c\u5bf9\u6570\u51e0\u7387\u7684\u5047\u8bbe\u3002\u56e0\u6b64\uff0c\u5982\u679c\u6570\u636e\u4e0d\u662f\u7ebf\u6027\u53ef\u5206\u7684\uff0c\u795e\u7ecf\u7f51\u7edc\u662f\u4e00\u4e2a\u66f4\u597d\u7684\u9009\u62e9\u3002\u4f46\u8fd9\u79cd\u7075\u6d3b\u6027\u662f\u4ee5\u96be\u4ee5\u89e3\u91ca\u53c2\u6570\u4f5c\u4e3a\u4ee3\u4ef7\u7684\u3002\u5176\u6240\u4ea7\u751f\u7684\u6a21\u578b\u662f\u901a\u8fc7\u6027\u80fd\u6307\u6807\u6765\u8bc4\u4f30\u7684\uff0c\u5982\u654f\u611f\u6027\u3001\u7279\u5f02\u6027\u3001\u51c6\u786e\u5ea6\u4ee5\u53ca\u63a5\u6536\u8005\u64cd\u4f5c\u7279\u5f81\u66f2\u7ebf\u4e0b\u7684\u9762\u79ef<\/span>\uff08\u89c1\u88684\uff09<\/span>\u3002\u795e\u7ecf\u7f51\u7edc\u5728\u8f93\u5165\u548c\u8f93\u51fa\u4e4b\u95f4\u5efa\u7acb\u4e86\u81f3\u5c11\u4e00\u4e2a\u9690\u85cf\u5c42\uff0c\u800c\u795e\u7ecf\u7f51\u7edc\u63d0\u9ad8\u6a21\u578b\u6027\u80fd\u7684\u597d\u5904\u5b9e\u9645\u4e0a\u6765\u81ea\u4e8e\u7b97\u6cd5\u4e2d\u591a\u4e2a\u9690\u85cf\u5c42\u7684\u80fd\u529b\u3002\u795e\u7ecf\u7f51\u7edc\u7684\u8bad\u7ec3\u8fc7\u7a0b\u4e3b\u8981\u5305\u62ec\u4e24\u4e2a\u6b65\u9aa4\uff1a\u9996\u5148\uff0c\u524d\u5411\u4f20\u64ad\u5c06\u8f93\u5165\u5c42\u6216\u5176\u4e4b\u540e\u9690\u85cf\u5c42\u7684\u6743\u91cd\u76f8\u7ed3\u5408\u3002\u800c\u540e\uff0c\u53cd\u5411\u4f20\u64ad\u5c06\u8f93\u51fa\u5c42\u6216\u5176\u4e4b\u524d\u7684\u9690\u85cf\u5c42\u5728\u5b9e\u9645\u503c\u548c\u9884\u6d4b\u503c\u7684\u8bef\u5dee\u57fa\u7840\u4e0a\u8fdb\u884c\u8c03\u6574\u3002\u901a\u8fc7\u8fd9\u79cd\u524d\u9988\u548c\u53cd\u5411\u4f20\u64ad\u7684\u8fed\u4ee3\uff0c\u795e\u7ecf\u7f51\u7edc\u8bad\u7ec3\u4ee5\u9002\u5e94\u4f20\u9012\u548c\u56de\u5f52\u53c2\u6570\u3002\u5982\u679c\u8bad\u7ec3\u540e\u6ca1\u6709\u8ba4\u771f\u5730\u6d4b\u8bd5\u548c\u4ea4\u53c9\u9a8c\u8bc1\uff0c\u90a3\u4e48\u795e\u7ecf\u7f51\u7edc\u5f88\u5bb9\u6613\u51fa\u73b0\u8fc7\u62df\u5408\u7684\u60c5\u51b5\u3002\u6b63\u5219\u5316\u53ef\u4ee5\u901a\u8fc7\u4ea4\u53c9\u9a8c\u8bc1\u6216bootstrapping\u6765\u89e3\u51b3\u8fd9\u4e2a\u95ee\u9898\uff0865\uff09\u3002<\/span><\/section>\n

<\/span><\/section>\n
\u53e6\u4e00\u79cd\u65b9\u6cd5\u662f\u4f7f\u7528\u8d1d\u53f6\u65af\u6846\u67b6\u3002\u8be5\u7b97\u6cd5\u7ed9\u51fa\u8ba1\u7b97\u53c2\u6570\u7684\u5206\u5e03\uff0c\u800c\u975e\u4e00\u4e2a\u70b9\u7684\u4f30\u8ba1\uff0c\u4ee5\u907f\u514d\u8fc7\u62df\u5408\u95ee\u9898\uff0866\uff09\u3002\u6b64\u5916\uff0c\u867d\u7136\u795e\u7ecf\u7f51\u7edc\u503e\u5411\u4e8e\u8fc7\u5ea6\u81ea\u4fe1\uff0c\u5373\u9884\u6d4b\u53ef\u80fd\u662f\u9519\u8bef\u7684\u5e76\u4e14\u5bb9\u6613\u53d7\u5230\u5bf9\u6297\u6027\u653b\u51fb(67)\uff0c\u800c\u8d1d\u53f6\u65af\u795e\u7ecf\u7f51\u7edc\uff0c\u5b83\u4ea7\u751f\u4e86\u4e00\u4e2a\u795e\u7ecf\u7f51\u7edc\u7684\u96c6\u5408\uff0c\u662f\u7a33\u5065\u548c\u51c6\u786e\u7684\uff0868\uff09\u3002\u8fd9\u5bf9\u4e8e\u63d0\u9ad8\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u7684\u4fe1\u4efb\u5ea6\u548c\u793e\u4f1a\u63a5\u53d7\u5ea6\u53ef\u80fd\u7279\u522b\u6709\u610f\u4e49\uff0c\u56e0\u4e3a\u5176\u8003\u8651\u4e86\u7b97\u6cd5\u548c\u7ed3\u679c\u6a21\u578b\u7684\u590d\u6742\u6027\u4e0e\u53ef\u89e3\u91ca\u6027\u7684\u5747\u8861\u3002<\/span><\/section>\n

<\/span><\/section>\n
\u672c\u6587\u63a5\u4e0b\u6765\u5c06\u4ecb\u7ecd\u4e0e\u9884\u6d4b\u6709\u5173\u7684\u7814\u7a76\u95ee\u9898\u4ee5\u53ca\u6765\u81ea\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u7684\u4f8b\u5b50\uff0c\u7279\u522b\u662f(i)\u793e\u4f1a\u6216\u5065\u5eb7\u7ed3\u679c\u7684\u9884\u6d4b\u3001(ii)\u8bc6\u522b\u548c\u8bc4\u4f30\u98ce\u9669\u56e0\u7d20\u4ee5\u53ca(iii)\u8bc6\u522b\u8fc7\u7a0b\u548c\u504f\u79bb\u201c\u6b63\u5e38\u201d\u7684\u60c5\u51b5\u3002<\/span><\/section>\n

<\/span><\/section>\n

\n
5.2 \u9884\u6d4b\u793e\u4f1a\u6216\u5065\u5eb7\u7684\u8f93\u51fa<\/span><\/strong><\/span><\/section>\n

<\/span><\/section>\n<\/h3>\n
\u5f88\u591a\u65f6\u5019\uff0c\u7406\u89e3\u4e00\u79cd\u793e\u4f1a\u73b0\u8c61\u6216\u75be\u75c5\u7684\u7b2c\u4e00\u6b65\u662f\u540c\u65f6\u8003\u8651\u5230\u793e\u4f1a\u4eba\u53e3\u3001\u5fc3\u7406\u3001\u5065\u5eb7\u3001\u6559\u80b2\u6216\u5176\u4ed6\u76f8\u5173\u7684\u51b3\u5b9a\u56e0\u7d20<\/span>\uff08\u9884\u6d4b\u56e0\u7d20\uff09<\/span>\uff0c\u5e76\u9884\u6d4b\u8c01\u4f1a\u663e\u793a\u51fa\u7ed3\u679c\u3002\u6211\u4eec\u9996\u5148\u8ba8\u8bba(i)\u8f93\u51fa\uff0c\u5373\u9884\u6d4b\u4e00\u4e2a\u901a\u5e38\u662f\u4e0d\u5229\u7684\u3001\u4f46\u6709\u65f6\u5bf9\u4e8e\u67d0\u4e9b\u7814\u7a76\u95ee\u9898\u4e5f\u662f\u6709\u76ca\u7684\u793e\u4f1a\u6216\u5065\u5eb7\u7ed3\u679c\u3002\u7136\u540e\uff0c\u8f6c\u5411\u5bfb\u627e (ii)\u6700\u5c0f\u6216\u6700\u4f73\u7684\u9884\u6d4b\u5668\u96c6\u4ee5\u53ca(iii)\u5bf9\u7f55\u89c1\u7ed3\u679c\u7684\u9884\u6d4b\u63d0\u51fa\u6539\u8fdb\u5efa\u8bae\u7b49\u7814\u7a76\u76ee\u6807\u3002<\/span><\/section>\n

<\/span><\/section>\n

\n
5.2.1 \u9884\u6d4b\u4e0d\u826f\u7ed3\u679c\u7684\u51c6\u786e\u6027<\/span><\/strong><\/section>\n

<\/span><\/section>\n<\/h4>\n
\u9884\u6d4b\u4e0d\u826f\u540e\u679c\u7684\u6a21\u578b\u5efa\u7acb\u901a\u5e38\u5305\u542b\u5728\u7814\u7a76\u4eba\u5458\u9009\u62e9\u7684\u5927\u578b\u7279\u5f81\u96c6\u4e0a\u8fdb\u884c\u8bad\u7ec3\u3002\u53ef\u80fd\u7684 \u6a21\u578b\u6027\u80fd\u7684\u6539\u8fdb\u5e94\u901a\u8fc7\u5efa\u7acb\u989d\u5916\u7684\u6a21\u578b\u3001\u6dfb\u52a0\u66f4\u591a\u7684\u4fe1\u606f<\/span>\uff08\u53d8\u91cf\uff09<\/span>\u3001\u6216\u8005\u8003\u8651\u9ad8\u9636\u76f8\u4e92\u4f5c\u7528\u7684\u6216\u975e\u7ebf\u6027\u7684\u6a21\u578b\u3002\u5178\u578b\u7684\u7814\u7a76\u95ee\u9898\u662f\u6d4b\u8bd5\u4e0e\u4f20\u7edf\u5efa\u6a21\u76f8\u6bd4\uff0cML\u5e26\u6765\u7684\u9884\u6d4b\u51c6\u786e\u6027\u6539\u5584\uff0c\u5982\u5173\u4e8e\u5065\u5eb7\u7684\u793e\u4f1a\u51b3\u5b9a\u56e0\u7d20\uff0858\uff09\u3001\u5bf9\u75f4\u5446\u75c7\u7684\u9884\u6d4b\uff0841\uff09\u4ee5\u53ca\u7cd6\u5c3f\u75c5\u548c\u7cd6\u5c3f\u75c5\u5e76\u53d1\u75c7\uff0869\uff09\u7684\u9884\u6d4b\u3002\u901a\u5e38\u57fa\u4e8eML\u7684\u6a21\u578b\u4e0d\u4f1a\u660e\u663e\u4f18\u4e8e\u4f20\u7edf\u7684\u57fa\u4e8e\u56de\u5f52\u7684\u6a21\u578b\u3002\u4f46\u662f\uff0c\u5176\u4ed6\u7814\u7a76\u8868\u660e\uff0c\u53ef\u4ee5\u7528\u957f\u77ed\u671f\u8bb0\u5fc6\u9012\u5f52\u795e\u7ecf\u7f51\u7edc\u901a\u8fc7\u7535\u5b50\u533b\u7597\u8bb0\u5f55\u6765\u4f30\u8ba1\u75c5\u4eba\u7684\u5269\u4f59\u9884\u671f\u5bff\u547d\u3002\u5269\u4f59\u5bff\u547d\u662f\u9762\u5411\u75c5\u4eba\u7684\u8861\u91cf\u6807\u51c6\uff0c\u53ef\u7528\u4e8e\u53ca\u65f6\u542f\u52a8\u9884\u5148\u62a4\u7406<\/span>\uff08\u4e34\u7ec8\uff09<\/span>\u8ba1\u5212\uff0870\uff09\u3002\u8fd9\u4e2a\u7814\u7a76\u7684\u4e00\u4e2a\u5f88\u597d\u7684\u4f8b\u8bc1\u662f\u5bf9\u9000\u4f0d\u519b\u4eba\u81ea\u6740\u7684\u9884\u6d4b\u3002\u5728\u4e0e\u5f53\u65f6\u6d3b\u7740\u76841%\u968f\u673a\u5339\u914d\u7684\u9000\u4f0d\u519b\u4eba\u6570\u636e\u96c6\u7684\u6d4b\u8bd5\u4e2d\uff0cBART\u662f\u51e0\u79cdML\u7b97\u6cd5\u4e2d\u6700\u597d\u7684\u4e00\u79cd\uff0871\uff09\u3002\u5176\u4ed6\u7684\u7814\u7a76\u4f7f\u7528\u4e86\u8d85\u7ea7\u5b66\u4e60\u5668\u8fdb\u884c\u6b7b\u4ea1\u98ce\u9669\u9884\u6d4b\uff0872\uff09\u3002\u4e00\u9879\u5173\u4e8e\u7528ML\u8fdb\u884c\u4e34\u5e8a\u98ce\u9669\u9884\u6d4b\u7684\u7cfb\u7edf\u53d1\u73b0\uff0cML\u65b9\u6cd5\u6bd4\u56de\u5f52\u7b97\u6cd5\u7684\u63d0\u5347\u4e0d\u660e\u663e\uff0c\u5e76\u4e14\u6279\u8bc4\u4e86\u8fc4\u4eca\u4e3a\u6b62\u6587\u732e\u4e2d\u7684\u4e00\u4e9b\u7f3a\u9677\uff0c\u7279\u522b\u662f\u7f3a\u4e4f\u6821\u51c6\uff0c\u5373\u6d4b\u8bd5\u9884\u6d4b\u7684\u53ef\u9760\u6027\uff0873\uff09\u3002\u7814\u7a76\u4eba\u5458\u53ef\u80fd\u610f\u8bc6\u5230\uff0c\u9884\u6d4b\u56e0\u7d20\u8d8a\u8fdc\uff0c\u8d8a\u96be\u8fbe\u5230\u7a33\u5065\u7684\u3001\u51c6\u786e\u7684\u9884\u6d4b\u7ed3\u679c\u3002\u4e3e\u4e00\u4e2a\u4f8b\u5b50\uff0c\u4e00\u9879\u7814\u7a76\u4f7f\u7528\u60e9\u6559\u8bbe\u65bd\u7684\u7279\u70b9\u548c\u56da\u72af\u7684\u603b\u4f53\u7279\u5f81\u6765\u9884\u6d4b\u76d1\u72f1\u66b4\u529b\u5e76\u7528\u56da\u72af\u95f4\u7684\u653b\u51fb\u6b21\u6570\u8fdb\u884c\u8bc4\u4f30\uff0c\u4f46\u8d85\u7ea7\u5b66\u4e60\u8005\u672a\u80fd\u8fbe\u5230\u5f88\u9ad8\u7684\u51c6\u786e\u7387\uff0874\uff09\u3002\u793e\u4f1a\u5730\u7406\u5b66\u4e2d\u7684\u7a7a\u95f4\u6620\u5c04\u540c\u6837\u88ab\u8bc1\u660e\u662f\u53ef\u7528\u7684\u3002\u4f8b\u5982\uff0c\u4e00\u9879\u7814\u7a76\u4f7f\u7528\u624b\u673a\u6570\u636e\u7ed8\u5236\u4e86\u8d2b\u56f0\u548c\u8d22\u5bcc\u7684\u5730\u7406\u5206\u5e03\u56fe\uff0c\u5e76\u5f97\u51fa\u4e86\u4e0e\u5b9e\u5730\u8c03\u67e5\u6570\u636e\u7c7b\u4f3c\u7684\u5206\u5e03\uff0875\uff09\u3002<\/span><\/section>\n

<\/span><\/section>\n
\u5728\u4e34\u5e8a\u5b9e\u8df5\u4e2d\uff0c\u76f8\u5173\u7684\u8f93\u51fa\u53ef\u80fd\u5e76\u4e0d\u662f\u4e0d\u5229\u7684\u5065\u5eb7\u7ed3\u679c\uff0c\u800c\u662f\u5e72\u9884\u7684\u5fc5\u8981\u6027\u548c\u6700\u4f73\u65f6\u673a\u3002\u5bf9\u4e34\u5e8a\u51b3\u7b56\u7684\u6700\u4f73\u65f6\u673a\u7684\u9884\u6d4b\u5df2\u7ecf\u88ab\u7814\u7a76\u8fc7\u4e86\u3002\u4f8b\u5982\uff0cvan der Schaar\u5b9e\u9a8c\u5ba4\u5df2\u7ecf\u5f00\u53d1\u4e86\u81ea\u52a8ML\u67b6\u6784\uff0c\u5e76\u79f0\u4e3aAutoPrognosis\uff0c\u4ee5\u9884\u6d4b\u4e0d\u5229\u7684\u5fc3\u8840\u7ba1\u75be\u75c5\u7ed3\u679c\uff0c\u4e14\u5176\u6027\u80fd\u4f18\u4e8e\u6bd4\u4f20\u7edf\u7684\u98ce\u9669\u8bc4\u5206(76)\u3002<\/span><\/section>\n

<\/span><\/section>\n

\n
5.2.2 \u793e\u4f1a\u6216\u5065\u5eb7\u7ed3\u679c\u7684\u6700\u5c0f\u548c\u6700\u4f73\u9884\u6d4b\u5668\u7ec4<\/span><\/strong><\/section>\n

<\/span><\/section>\n<\/h4>\n
\u793e\u4f1a\u6216\u5065\u5eb7\u7684\u8f93\u51fa\u4e5f\u53ef\u80fd\u662f\u5e0c\u671b\u627e\u5230\u4e00\u4e2a\u6700\u5c0f\u6216\u6700\u4f73<\/span>\uff08\u6700\u7b80\u660e\uff09<\/span>\u7684\u9884\u6d4b\u5668\u96c6\u5408\u3002\u56de\u5230\u524d\u9762\u63d0\u5230\u7684\u9884\u6d4b\u975e\u4f20\u67d3\u6027\u75be\u75c5\u7684\u4eba\u53e3\u6c34\u5e73\u6d41\u884c\u7387\u7684\u4f8b\u5b50\uff0c\u8fd9\u4e9b\u75be\u75c5\u662f\u7528\u6700\u5c0f\u7684\u793e\u4f1a\u4eba\u53e3\u5b66\u9884\u6d4b\u5668\u7ec4\u6765\u4f30\u8ba1\u7684\uff0844\uff09\u3002\u5728\u75f4\u5446\u75c7\u98ce\u9669\u9884\u6d4b\u65b9\u9762\uff0c\u81f3\u4eca\u4ecd\u6ca1\u6709\u5f3a\u6709\u529b\u7684\u7b97\u6cd5\u3002\u6700\u8fd1\u7528LASSO\u5f00\u53d1\u4e86\u4e00\u4e2a\u9884\u6d4b10\u5e74\u4ee5\u4e0a\u75f4\u5446\u75c7\u7684\u6700\u4f73\u6a21\u578b\uff0877\uff09\u3002\u53e6\u4e00\u9879\u7814\u7a76\u7ec4\u5408LASSO\u548c\u968f\u673a\u68ee\u6797\u7b97\u6cd5\uff0c\u4ee5\u5bfb\u6c42\u4e00\u4e2a\u6700\u4f73\u7684\u9884\u6d4b\u5668\u7ec4\u6765\u89e3\u91ca\u67aa\u652f\u66b4\u529b\u7684\u5dee\u5f02\uff0878\uff09\u3002<\/span><\/section>\n

\n

<\/section>\n
5.2.3 \u7f55\u89c1\u7ed3\u679c\u9884\u6d4b<\/span><\/strong><\/section>\n

<\/span><\/section>\n<\/h4>\n
\u5065\u5eb7\u548c\u793e\u4f1a\u79d1\u5b66\u4e2d\u7684\u4e00\u4e2a\u5178\u578b\u95ee\u9898\u662f\u9884\u6d4b\u7f55\u89c1\u7684\u4e8b\u4ef6\uff0c\u5982\u75be\u75c5\u3001\u72af\u7f6a\u3001\u5b66\u4e60\u56f0\u96be\u548c\u79bb\u5a5a\u3002\u5728\u8fd9\u79cd\u60c5\u51b5\u4e0b\uff0c\u53ea\u6709\u975e\u5e38\u5c0f\u6bd4\u4f8b\u7684\u88ab\u89c2\u5bdf\u4eba\u53e3\u4f1a\u51fa\u73b0\u611f\u5174\u8da3\u7684\u7ed3\u679c\u3002\u4f8b\u5982\uff0c\u4e0e\u603b\u4eba\u53e3\u76f8\u6bd4\uff0c\u72af\u7f6a\u8005\u7684\u5360\u6bd4\u975e\u5e38\u5c0f\uff1b\u81ea\u95ed\u75c7\u53ea\u57282%\u5de6\u53f3\u7684\u513f\u7ae5\u4e2d\u6d41\u884c\uff1b\u591a\u8fbe\u4e09\u5206\u4e4b\u4e00\u7684\u5df2\u5a5a\u4eba\u58eb\u4f1a\u63d0\u51fa\u79bb\u5a5a\u3002\u901a\u8fc7ML\u6765\u9884\u6d4b\u7f55\u89c1\u7684\u7ed3\u679c<\/span>\uff08\u800c\u7f55\u89c1\u53ef\u4ee5\u5b9a\u4e49\u4e3a\u4efb\u4f55\u6982\u7387\u4f4e\u4e8e50%\u7684\u6848\u4f8b\uff09<\/span>\uff0c\u5206\u7c7b\u7b97\u6cd5\u901a\u5e38\u4f1a\u7b80\u5355\u5730\u5f00\u53d1\u4e00\u4e2a\u6a21\u578b\uff0c\u8be5\u6a21\u578b\u53ea\u9884\u6d4b\u4e0d\u53d1\u751f\u7684\u7ed3\u679c\uff0c\u56e0\u4e3a\u7b97\u6cd5\u4f1a\u68c0\u6d4b\u5230\u5728\u5927\u591a\u6570\u60c5\u51b5\u4e0b\uff0c\u201c0\u201d\u7684\u731c\u6d4b\u662f\u6b63\u786e\u7684\u3002\u7814\u7a76\u4eba\u5458\u53ef\u4ee5\u901a\u8fc7\u51e0\u79cd\u7b56\u7565\u89e3\u51b3\u8fd9\u4e2a\u95ee\u9898\u3002<\/span><\/section>\n

<\/span><\/section>\n
1\uff09\u91cd\u65b0\u5b9a\u4e49\u7ed3\u679c\u3002\u4f8b\u5982\uff0c\u8f6c\u5316\u4e3a\u4e00\u4e2a\u56de\u5f52\u95ee\u9898\u800c\u4e0d\u662f\u5206\u7c7b\u95ee\u9898\uff0c\u4f8b\u5982\uff0c\u4f7f\u7528\u6291\u90c1\u91cf\u8868\u7684\u5168\u90e8\u8303\u56f4\u4f5c\u4e3a\u8fde\u7eed\u6216\u6cca\u677e\u5206\u5e03\u7684\u53d8\u91cf\u3002\u800c\u4e0d\u662f\u4e2d\u5ea6\/\u91cd\u5ea6\u6291\u90c1\u75c7\u4e34\u5e8a\u76f8\u5173\u7684\u4e8c\u5143<\/span>\uff08\u548c\u7f55\u89c1\u7684\uff09<\/span>\u5206\u7c7b\u3002<\/span><\/section>\n

<\/span><\/section>\n
2\uff09\u901a\u8fc7\u5bf9\u5177\u6709\u7f55\u89c1\u7ed3\u679c\u7684\u7fa4\u4f53\u8fdb\u884c\u8d85\u989d\u62bd\u6837\u6765\u6539\u53d8\u6837\u672c\u4e2d\u7684\u7ed3\u679c\u5206\u5e03\uff0c\u4f8b\u5982\u5408\u6210\u5c11\u6570\u4eba\u8d85\u989d\u62bd\u6837\u6280\u672f(79)\u662f\u5f88\u597d\u7684\u65b9\u6cd5\u3002\u540c\u6837\u5730\uff0c\u5bf9\u6ca1\u6709\u7ed3\u679c\u7684\u7fa4\u4f53\u8fdb\u884c\u4e0b\u91c7\u6837<\/span>\uff08\u6b20\u91c7\u6837\uff09<\/span>\u4e5f\u662f\u53ef\u4ee5\u7684\u3002\u4e3e\u4e00\u4e2a\u9ad8\u5ea6\u4e0d\u5e73\u8861\u7684\u91d1\u878d\u670d\u52a1\u6d88\u8d39\u8005\u884c\u4e3a\u7684\u4f8b\u5b50\uff0c\u62b5\u62bc\u8d37\u6b3e\u8fdd\u7ea6\u7684\u9884\u6d4b\u548c\u9ad8\u98ce\u9669\u6d88\u8d39\u8005\u7684\u8bc6\u522b\u5206\u522b\u5bf9\u94f6\u884c\u548c\u8d37\u6b3e\u516c\u53f8\u6765\u8bf4\u662f\u81f3\u5173\u91cd\u8981\u7684\u3002\u6700\u8fd1\u7684\u4e00\u9879\u6d88\u8d39\u8005\u4ea4\u6613\u6570\u636e\u7684\u7814\u7a76\u5bf9\u672a\u8fdd\u7ea6\u7684\u5ba2\u6237\u8fdb\u884c\u4e86\u62bd\u6837\u8c03\u67e5\uff0c\u5e76\u8bad\u7ec3\u4e86\u4e0e\u968f\u673a\u68ee\u6797\u7ed3\u5408\u7684\u5377\u79ef\u795e\u7ecf\u7f51\u7edc\uff0c\u8be5\u7f51\u7edc\u7ed3\u6784\u5e26\u6765\u4e86\u6700\u9ad8\u7684\u9884\u6d4b\u51c6\u786e\u6027(80)\u3002\u4e00\u4e9b\u7814\u7a76\u62a5\u544a\u79f0\uff0c\u901a\u8fc7\u5bf9\u9ad8\u5ea6\u4e0d\u5e73\u8861\u7684\u6570\u636e\u8fdb\u884c\u91cd\u590d\u7684\u968f\u673a\u5b50\u62bd\u6837\u53ef\u4ee5\u4ee5\u9884\u6d4b\u75be\u75c5\uff0c\u5e76\u4e14\u968f\u673a\u68ee\u6797\u7684\u8868\u73b0\u4f18\u4e8e\u652f\u6301\u5411\u91cf\u673a\u3001\u888b\u6cd5\u4ee5\u53ca\u5176\u4ed6\u65b9\u6cd5\uff0881\uff09\u3002\u5e94\u8be5\u6ce8\u610f\u7684\u662f\uff0c\u5728\u4ea4\u53c9\u9a8c\u8bc1\u7684\u60c5\u51b5\u4e0b\uff0c\u91c7\u6837\u7b56\u7565\u5e94\u8be5\u88ab\u591a\u6b21\u91cd\u590d\u3002\u4ece\u66f4\u5e7f\u6cdb\u7684\u610f\u4e49\u4e0a\u8bb2\uff0c\u9700\u8981\u8fdb\u884c\u654f\u611f\u6027\u68c0\u67e5\uff0c\u4ee5\u4fbf\u7814\u7a76\u4e0d\u540c\u7684\u62bd\u6837\u7b56\u7565\u5bf9\u6027\u80fd\u548c\u53ef\u63a8\u5e7f\u6027\u7684\u5f71\u54cd\u3002<\/span><\/section>\n

<\/span><\/section>\n
3\uff09\u4e00\u4e2a\u7f55\u89c1\u7684\u7ed3\u679c\u5728\u9884\u9009\u7684\u6837\u672c\u4e2d\u53ef\u80fd\u66f4\u9891\u7e41\u3002\u4f8b\u5982\uff0c\u75f4\u5446\u75c7\u7814\u7a76\u7684\u90e8\u5206\u91cd\u70b9\u662f\u63d0\u9ad8\u4ece\u8f7b\u5ea6\u8ba4\u77e5\u969c\u788d<\/span>\uff08MCI\uff09<\/span>\u8f6c\u5316\u4e3a\u75f4\u5446\u75c7\u7684\u9884\u6d4b\u7387\u3002\u53ea\u6709\u5206\u6790\u5f97\u5230\u7684\u7ed3\u8bba\u4e0d\u9700\u8981\u5728\u6574\u4e2a\u4eba\u7fa4\u4e2d\u6210\u7acb\u65f6\uff0c\u6837\u672c\u9009\u62e9\u624d\u80fd\u6210\u529f\uff0c\u5426\u5219\u8fd9\u79cd\u7b56\u7565\u4f1a\u5f15\u5165\u504f\u89c1<\/span>\uff08\u5bf9\u7ed3\u679c\u7684\u6761\u4ef6\uff09<\/span>\u3002MCI\u60a3\u8005\u7684\u9ad8\u5371\u6837\u672c\u53ef\u4ee5\u7528\u6765\u8bad\u7ec3\u4e00\u4e2a\u6a21\u578b\uff0c\u6839\u636e\u95ee\u5377\u8c03\u67e5\u3001\u751f\u7269\u6807\u5fd7\u7269\u6216\u57fa\u4e8e\u5f71\u50cf\u7684\u53d8\u91cf\u6765\u533a\u5206\u8f6c\u6362\u8005\u548c\u975e\u8f6c\u6362\u8005\u3002\u5728\u4f17\u591a\u7684\u4f8b\u5b50\u4e2d\uff0c\u4e00\u4e9b\u7814\u7a76\u7528\u4e0d\u540c\u7684ML\u7b97\u6cd5\u6d4b\u8bd5\u4e86\u8ba4\u77e5\u50a8\u5907\u5bf9\u9884\u6d4b\u8f6c\u6362\u4e3a\u75f4\u5446\u7684\u4f5c\u7528\uff0882\uff09\uff0c\u5e76\u5728\u4ee5\u524d\u672a\u89c1\u8fc7\u7684\u5bf9\u8c61\u4e2d\u6d4b\u8bd5\u4e86\u4ed6\u4eec\u7528\u652f\u6301\u5411\u91cf\u673a\u7b97\u6cd5\u5f00\u53d1\u7684\u6a21\u578b\uff0883\uff09\u3002\u867d\u7136\u75f4\u5446\u98ce\u9669\u9884\u6d4b\u6a21\u578b\u6709\u4e86\u5f88\u5927\u7684\u6539\u5584\uff0c\u7279\u522b\u662f\u5728\u5411\u6a21\u578b\u4e2d\u52a0\u5165\u9057\u4f20\u548c\u6210\u50cf\u4fe1\u606f\u540e\uff0c\u4f46\u5728\u9884\u6d4b\u9ad8\u5371\u6837\u672c\u7684\u8ba4\u77e5\u6027\u80fd\u6d4b\u8bd5\u4e0b\u964d\u7684\u80fd\u529b\u65b9\u9762\u6ca1\u6709\u660e\u663e\u8fdb\u5c55\uff0884\uff09<\/span><\/section>\n

<\/span><\/section>\n
4\uff09\u6a21\u62df\u6570\u636e\u96c6\u548c “\u865a\u62df\u961f\u5217 “\u5728\u67d0\u4e9b\u60c5\u51b5\u4e0b\u53ef\u80fd\u662f\u6709\u7528\u7684\u3002<\/span><\/section>\n

<\/span><\/section>\n

\n
5.3 \u65b0\u7684\u6216\u5df2\u77e5\u98ce\u9669\u56e0\u7d20\u8bc6\u522b\u548c\u8bc4\u4f30<\/span><\/strong><\/span><\/section>\n

<\/span><\/section>\n<\/h3>\n
\u51fa\u4e8e\u5bf9\u8f93\u5165\u7684\u5174\u8da3\uff0c\u5373\u793e\u4f1a\u6216\u5065\u5eb7\u7ed3\u679c\u7684\u9884\u6d4b\u56e0\u7d20\uff0c\u4e00\u4e9b\u7814\u7a76\u5229\u7528\u5927\u578b\u9884\u6d4b\u5668\u96c6\u6765\u786e\u5b9a \u8bc6\u522b\u4ee5\u524d\u672a\u77e5\u7684\u793e\u4f1a\u6216\u5065\u5eb7\u7ed3\u679c\u7684\u9884\u6d4b\u56e0\u7d20\uff0c\u6216\u5728\u5176\u4ed6\u53d8\u91cf\u7684\u80cc\u666f\u4e0b\u8bc4\u4f30\u5176\u9884\u6d4b\u80fd\u529b\u3002\u9664\u4e86\u9700\u8981\u8003\u8651\u7ef4\u5ea6\u8bc5\u5492\u5916\uff0c\u91c7\u7528\u51e0\u79cdML\u65b9\u6cd5\u6d4b\u8bd5\u4ee5\u524d\u672a\u6d4b\u8bd5\u8fc7\u7684\u9884\u6d4b\u56e0\u5b50\u96c6\u53ef\u6709\u52a9\u4e8e\u5e73\u8861\u5176\u5c40\u9650\u6027\u3002\u4ee5\u6b64\u4e3a\u76ee\u7684\u5f00\u5c55\u4e86\u4e00\u7cfb\u5217\u7814\u7a76\uff0c\u4f8b\u5982\u6d4b\u8bd5\u4e0e\u513f\u7ae5\u8ba4\u77e5\u8868\u73b0\u6709\u5173\u7684\u5019\u9009\u53ef\u6539\u53d8\u56e0\u7d20\uff0885\uff09\u3002\u5728\u53e6\u4e00\u9879\u5e94\u7528\u4e2d\uff0c\u68af\u5ea6\u63d0\u5347\u548cSHAP\u56fe\u88ab\u7528\u6765\u8bc6\u522b\u9752\u5c11\u5e74\u81ea\u6740\u60f3\u6cd5\u548c\u884c\u4e3a\u7684\u5341\u5927\u98ce\u9669\u56e0\u7d20\uff0c\u7ed3\u679c\u8868\u660e\u8fd9\u4e9b\u56e0\u7d20\u90fd\u4e0e\u793e\u4f1a\u4eba\u53e3\u56e0\u7d20\u4ee5\u53ca\u5bb6\u5ead\u548c\u540c\u4f34\u5173\u7cfb\u6709\u5173\uff0886\uff09\u3002<\/span><\/section>\n

<\/span><\/section>\n
\u4e00\u9879\u7814\u7a76\u901a\u8fc7\u53ef\u7a7f\u6234\u8bbe\u5907\u8fdb\u884c\u7684\u8ff7\u4f60\u7cbe\u795e\u72b6\u6001\u68c0\u67e5<\/span>\uff08MMSE\uff09<\/span>\u7684\u5206\u6570\uff0c\u63a2\u7a76\u8ba4\u77e5\u529f\u80fd\u4e0e\u5df2\u77e5\u751f\u6d3b\u65b9\u5f0f\u7684\u5173\u8054\uff0887\uff09\u3002\u867d\u7136\u7531\u4e8e\u7814\u7a76\u7684\u6a2a\u65ad\u9762\u6027\u8d28\uff0c\u65e0\u6cd5\u786e\u5b9a\u56e0\u679c\u5173\u7cfb\uff0c\u4f46\u90e8\u5206\u4f9d\u8d56\u56fe\u7684\u53ef\u89c6\u5316\u663e\u793a\u4e86\u975e\u7ebf\u6027\u5173\u8054\uff0c\u5982\u5173\u8054\u5728\u67d0\u4e2a\u9608\u503c\u540e\u8d8b\u4e8e\u5e73\u7a33\uff0c\u6216\u4e0e\u53cdU\u578b\u5173\u7cfb\u76f8\u5173\uff0887\uff09\u3002\u8fd9\u6837\u7684\u7814\u7a76\u6709\u52a9\u4e8e\u6539\u5584\u5bf9\u66b4\u9732-\u7ed3\u679c\u5173\u8054\u4e2d\u7684\u9884\u671f\u548c\u5b9e\u9645\u5173\u7cfb\u7684\u601d\u8003\uff0c\u5373\u671f\u671b\u4ece\u67d0\u79cd\u751f\u6d3b\u65b9\u5f0f\u4e2d\u83b7\u5f97\u7684\u76ca\u5904\u662f\u6301\u7eed\u589e\u52a0\u7684<\/span>\uff08\u7ebf\u6027\u6216\u4e8c\u6b21\u5242\u91cf-\u53cd\u5e94\u5173\u7cfb\uff09<\/span>\uff0c\u6216\u8005\u5728\u67d0\u4e00\u6570\u503c\u540e\u8d8b\u4e8e\u5e73\u7f13\uff0c\u6b64\u540e\u4e0d\u80fd\u518d\u6709\u4efb\u4f55\u6539\u5584<\/span>\uff08\u9608\u503c\u6a21\u578b\uff09<\/span>\uff1f\u5bf9\u8fd9\u4e9b\u8c03\u67e5\u6765\u8bf4\uff0c\u5bf9\u66b4\u9732\u548c\u7ed3\u679c\u8fdb\u884c\u7cbe\u7ec6\u7684\u8bc4\u4f30\u662f\u5fc5\u8981\u7684\u3002\u4e00\u9879\u7814\u7a76\u7528LASSO\u9884\u6d4b\u963f\u7247\u7c7b\u836f\u7269\u8fc7\u91cf\u7684\u6b7b\u4ea1\u7387\uff0c\u53d1\u73b0\u4ee5\u524d\u6ca1\u6709\u53d1\u73b0\u7684\u76f8\u5173\u90bb\u91cc\u7279\u5f81\uff0c\u5982\u5c45\u4f4f\u7a33\u5b9a\u6027\u3001\u79cd\u65cf\/\u6c11\u65cf\u5206\u5e03\u548c\u793e\u4f1a\u9694\u79bb\uff0888\uff09\u3002\u53e6\u4e00\u9879\u7814\u7a76\u5728\u6a2a\u65ad\u9762\u8bbe\u8ba1\u4e2d\u6d4b\u8bd5\u4e86\u7ae5\u5e74\u4e0d\u826f\u7ecf\u5386\u4e0e\u667a\u529b\u7684\u5173\u8054\uff0889\uff09\u3002\u4ece\u6240\u9009\u62e9\u7684\u672f\u8bed\u201c\u98ce\u9669\u56e0\u7d20\u201d\u5230\u66f4\u4e2d\u6027\u7684\u201c\u51b3\u5b9a\u56e0\u7d20\u201d\u6216 \u201c\u9884\u6d4b\u56e0\u7d20\u201d\uff0c\u53e6\u4e00\u9879\u7814\u7a76\u65e8\u5728\u901a\u8fc7\u6df1\u5ea6<\/span>\uff08\u4eba\u5de5\uff09<\/span>\u795e\u7ecf\u7f51\u7edc\u68c0\u9a8c\u7ecf\u5178\u7684\u51b3\u7b56\u7406\u8bba\uff0890\uff09\uff0c\u5229\u7528\u5927\u89c4\u6a21\u7684\u98ce\u9669\u9009\u62e9\u5b9e\u9a8c\uff0c\u786e\u5b9a\u4eba\u7c7b\u51b3\u7b56\u4e2d\u7684\u9a71\u52a8\u56e0\u7d20<\/span>\uff08\u9884\u6d4b\u56e0\u7d20\uff09<\/span>\u3002<\/span><\/section>\n

<\/span><\/section>\n

\n
5.4 \u8bc6\u522b\u6b63\u5e38\u8fc7\u7a0b\u53ca\u5176\u6ce2\u52a8<\/span><\/strong><\/span><\/section>\n

<\/span><\/section>\n<\/h3>\n
\u5728\u8c03\u67e5\u793e\u4f1a\u548c\u5065\u5eb7\u7ed3\u679c\u7684\u8f68\u8ff9\u65f6\uff0c\u7814\u7a76\u4eba\u5458\u53ef\u80fd\u5e0c\u671b\u4e0d\u4ec5\u4ec5\u662f\u63cf\u8ff0\u6027\u7814\u7a76\uff0c\u5982\u679c\u4e0d\u540c\u7684\u72b6\u6001\u6216\u8f68\u8ff9\u5df2\u7ecf\u88ab\u4e3b\u9898\u6240\u5b9a\u4e49\uff0c\u4f8b\u5982\u75be\u75c5\u7684\u4e25\u91cd\u7a0b\u5ea6\u6216\u6559\u80b2\u6216\u804c\u4e1a\u6c34\u5e73\uff0c\u5219\u5e94\u91c7\u7528\u9884\u6d4b\u6027\u773c\u5149\u3002\u4e00\u9879\u5065\u5eb7\u7d22\u8d54\u6570\u636e\u7684\u7814\u7a76\u91c7\u7528\u4e86\u68af\u5ea6\u63d0\u5347\u548c\u79fb\u52a8\u65f6\u95f4\u7a97\u53e3\u65b9\u6cd5\uff0c\u8c03\u67e5\u4e86\u6700\u4e25\u91cd\u7684\u6162\u6027\u963b\u585e\u6027\u80ba\u75c5\u548c\u5e38\u89c1\u75be\u75c5\u8f68\u8ff9\u7684\u9884\u6d4b\u56e0\u7d20\u3002\u7814\u7a76\u5458\u8bbe\u8ba1\u4e86\u4e00\u4e9b\u8bca\u65ad<\/span>\uff08\u5982\u547c\u5438\u8870\u7aed\uff09<\/span>\u3001\u836f\u7269<\/span>\uff08\u5982\u6297\u80c6\u78b1\u836f\u7269\uff09<\/span>\u4ee5\u53ca\u4e0e\u968f\u540e\u7684\u6700\u4e25\u91cd\u7684\u6162\u6027\u963b\u585e\u6027\u80ba\u75c5\u8bca\u65ad\u6709\u5173\u7684\u7a0b\u5e8f\uff0891\uff09\u3002\u672c\u7814\u7a76\u68c0\u6d4b\u5230\u7684\u65f6\u95f4\u6a21\u5f0f\u66f4\u591a\u7684\u662f\u4ee3\u8868\u4e86\u4e0e\u4fdd\u5065\u6709\u5173\u7684\u75be\u75c5\u987a\u5e8f\uff0c\u800c\u4e0d\u5e94\u8be5\u5728\u56e0\u679c\u5173\u7cfb\u7684\u610f\u4e49\u4e0a\u8fdb\u884c\u89e3\u91ca\uff0891\uff09\u3002\u5728\u5176\u4ed6\u60c5\u51b5\u4e0b\uff0c\u68c0\u6d4b\u5230\u7684\u65f6\u95f4\u6a21\u5f0f\u53ef\u80fd\u66f4\u52a0\u7a33\u5065\u3002\u5bf9\u8001\u9f84\u5316\u8fc7\u7a0b\u611f\u5174\u8da3\u7684\u7814\u7a76\u4eba\u5458\u5e0c\u671b\u786e\u5b9a\u4e0e\u8001\u9f84\u5316\u76f8\u5173\u7684\u793e\u4f1a\u6216\u5065\u5eb7\u7ed3\u679c\u7684\u6b63\u5e38\u8f68\u8ff9\uff0c\u5e76\u901a\u8fc7\u68af\u5ea6\u63d0\u5347\u7b49\u65b9\u6cd5\u6765\u786e\u5b9a\u4e0e\u6b63\u5e38\u8f68\u8ff9\u7684\u504f\u5dee\uff0c\u8fd9\u53ef\u5f52\u7eb3\u4e3a\u4e00\u4e2a\u9884\u6d4b\u95ee\u9898\uff0892\uff09\u3002\u7814\u7a76\u4eba\u5458\u5e94\u8be5\u610f\u8bc6\u5230\uff0c\u56e0\u4e3a\u5b9a\u4e49\u6b63\u5e38\u7684\u8870\u8001\u8f68\u8ff9\u5e76\u4e0d\u7b80\u5355\uff0c\u8fd9\u4e2a\u76ee\u6807\u5bf9\u6570\u636e\u63d0\u51fa\u4e86\u6781\u9ad8\u7684\u8981\u6c42\u3002\u7406\u60f3\u60c5\u51b5\u4e0b\uff0c\u9700\u8981\u63d0\u4f9b\u8db3\u591f\u7684\u4fe1\u606f\u4ee5\u786e\u4fdd\u53ef\u89e3\u91ca\u6027\u548c\u53ef\u590d\u73b0\u6027\u3002\u4e0b\u6587\u5c06\u4ee5\u56e0\u679c\u7814\u7a76\u4e3a\u76ee\u6807\u63cf\u8ff0\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u7684\u7814\u7a76\u3002<\/span><\/section>\n

<\/span><\/section>\n

\n

<\/section>\n<\/h2>\n

\n
\n
\n
\n
\n

<\/section>\n

<\/section>\n
\n
<\/svg><\/section>\n<\/section>\n<\/section>\n
\n

6. \u56e0\u679c\u63a8\u65ad\u4e2d\u7684\u673a\u5668\u5b66\u4e60<\/strong><\/strong><\/p>\n<\/section>\n

\n

<\/section>\n

<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/h3>\n

<\/section>\n
\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u7814\u7a76\u8005\u6240\u8ffd\u6c42\u7684\u5927\u90e8\u5206\u90fd\u4e0e\u5bfb\u627e\u4e16\u754c\u4e0a\u67d0\u4e00\u7279\u5f81\u7684\u539f\u56e0\u6216\u540e\u679c\u6709\u5173\uff0c\u6545\u5f80\u5f80\u4f1a\u5bfb\u6c42\u56e0\u679c\u95ee\u9898\u7684\u7b54\u6848\u3002\u5982\u679c\u4e0d\u4ec5\u8981\u627e\u51fa\u9884\u6d4b\u56e0\u7d20\uff0c\u8fd8\u8981\u627e\u51fa\u98ce\u9669\u6216\u4fdd\u62a4\u56e0\u7d20\uff0c\u4f8b\u5982\uff0c\u7528\u4e8e\u67d0\u79cd\u4e0d\u5229\u7684\u793e\u4f1a\u6216\u5065\u5eb7\u7ed3\u679c\u7684\u9884\u9632\u3002\u5982\u679c\u4e0d\u4ec5\u60f3\u4e86\u89e3\u51b3\u5b9a\u56e0\u7d20\uff0c\u8fd8\u60f3\u8fdb\u884c\u5e72\u9884\uff0c\u90a3\u4e48\u5c31\u9700\u8981\u4e86\u89e3\u75be\u75c5\u7684\u56e0\u679c\u51b3\u5b9a\u56e0\u7d20\u3002<\/span><\/section>\n

<\/span><\/section>\n
\u5728\u5bf9\u56e0\u679c\u5173\u7cfb\u95ee\u9898\u627e\u5230\u6ee1\u610f\u7684\u7b54\u6848\u4e4b\u524d\uff0c\u9700\u6ee1\u8db3\u7edf\u8ba1\u5206\u6790\u7684\u591a\u79cd\u8981\u6c42\u3002\u7528\u4e8e\u56e0\u679c\u63a8\u65ad\u7684ML\u9700\u8981\u9886\u57df\u77e5\u8bc6\uff0c\u5373\u4e3b\u9898\u77e5\u8bc6\u3002\u6839\u636e\u53d8\u91cf\u5728\u63cf\u8ff0\u53d8\u91cf\u4e4b\u95f4\u5047\u5b9a\u56e0\u679c\u5173\u7cfb\u7684\u6709\u5411\u65e0\u73af\u56fe<\/span>\uff08DAG\uff09<\/span>\u4e0a\u7684\u4f4d\u7f6e\uff0c\u5408\u9002\u5730\u9009\u62e9\u53d8\u91cf\u662f\u81f3\u5173\u91cd\u8981\u7684\uff0893\uff0c94\uff09\uff0c\u8fd9\u5728\u5065\u5eb7\u7814\u7a76\u4e2d\u7684\u5e94\u7528\u8d8a\u6765\u8d8a\u591a\uff0895\uff09\u3002\u4efb\u4f55\u65e8\u5728\u8fdb\u884c\u56e0\u679c\u63a8\u65ad\u7684\u7edf\u8ba1\u5206\u6790\u901a\u5e38\u4f1a\u9009\u62e9\u90a3\u4e9b\u53ef\u4ee5\u6ee1\u8db3\u56e0\u679c\u63a8\u65ad\u5047\u8bbe\u7684\u6570\u636e\u96c6\uff1a\u53ef\u4ea4\u6362\u6027<\/span>\uff08\u65e0\u77e5\u6027\uff09<\/span>\uff0c\u5373\u5bf9\u4e8e\u6240\u6709\u6ca1\u6709\u63a5\u53d7\u7279\u5b9a\u6cbb\u7597\u7684\u4eba\u6765\u8bf4\uff0c\u5982\u679c\u4ed6\u4eec\u63a5\u53d7\u4e86\u6cbb\u7597\uff0c\u5176\u7ed3\u679c\u5c06\u4e0e\u63a5\u53d7\u6cbb\u7597\u7684\u4eba\u76f8\u540c<\/span>\uff08\u7ed3\u679c\u7684\u53cd\u4e8b\u5b9e\u6982\u7387\uff09<\/span>\uff1b\u79ef\u6781\u6027\uff0c\u5373\u66b4\u9732\u548c\u6df7\u6742\u56e0\u7d20\u7684\u6bcf\u4e00\u79cd\u503c\u7684\u6240\u6709\u53ef\u80fd\u503c\u90fd\u662f\u53ef\u7528\u7684\uff0c\u6216\u8005\u5728\u6570\u636e\u96c6\u4e2d\u5df2\u7ecf\u88ab\u5206\u914d\uff1b\u4e00\u81f4\u6027\uff0c\u5373\u4e00\u4e2a\u4eba\u5728\u5176\u89c2\u5bdf\u5230\u7684\u66b4\u9732\u5386\u53f2\u4e0b\u7684\u6f5c\u5728\u7ed3\u679c\u6b63\u662f\u4ed6\u4eec\u7684\u89c2\u5bdf\u7ed3\u679c\uff0896\uff09\u3002\u7136\u800c\uff0c\u5373\u4fbf\u5728\u8fdd\u53cd\u8fd9\u4e9b\u6761\u4ef6<\/span>\uff08\u6709\u9650\u7684\uff09<\/span>\u7684\u60c5\u51b5\u4e0b\uff0c\u5982\u679c\u968f\u673a\u5bf9\u7167\u8bd5\u9a8c\u4e0d\u6ee1\u8db3\u4f26\u7406\uff0c\u6216\u7531\u4e8e\u5176\u4ed6\u539f\u56e0\u4e0d\u53ef\u884c\uff0c\u90a3\u4e48\u901a\u8fc7\u5b9a\u4e49\u5047\u8bbe\u7684\u5e72\u9884\u63aa\u65bd\u6216\u201c\u76ee\u6807\u8bd5\u9a8c\u201d\u4e5f\u53ef\u4ee5\u53d6\u5f97\u4e00\u4e9b\u8fdb\u5c55\u3002\u901a\u8fc7\u4f7f\u7528\u5177\u6709\u6f5c\u5728\u7ed3\u679c\u6846\u67b6\u7684\u89c2\u5bdf\u6027\u6570\u636e\uff0c\u6211\u4eec\u53ef\u4ee5\u6a21\u4eff\u76ee\u6807\u8bd5\u9a8c\uff0897, 98\uff09\u3002\u6700\u8fd1\u7684\u4e00\u9879\u6a21\u4eff\u76ee\u6807\u8bd5\u9a8c\u7684\u7814\u7a76\u7528\u76ee\u6807\u6700\u5927\u4f3c\u7136\u4f30\u8ba1<\/span>\uff08TMLE\uff09<\/span>\u6765\u4f30\u8ba1\u7cfb\u6570\uff0c\u5e76\u6d4b\u8bd5\u5e72\u9884\u53ef\u53d8\u56e0\u7d20\u5bf9\u513f\u7ae5\u671fBMI\u53cd\u5f39\u7684\u5f71\u54cd\uff0899\uff09\u3002<\/span><\/section>\n

<\/span><\/section>\n
\u6839\u636eHern\u00e1n\u7b49\u4eba\uff086\uff09\u7684\u5206\u7c7b\uff0c\u9996\u5148\u5728\u6b64\u8003\u8651\u7528\u4e8e\u53cd\u4e8b\u5b9e\u9884\u6d4b\u7684ML\uff0c\u4f46\u672c\u6587\u540c\u65f6\u8003\u8651\u7528\u4e8e\u56e0\u679c\u53d1\u73b0\u7684ML\u4ee5\u6269\u5c55\u8be5\u6846\u67b6\uff0c\u5373\u53d8\u91cf\u4e4b\u95f4\u7684\u56e0\u679c\u7ed3\u6784\u662f\u4ece\u6570\u636e\u4e2d\u5b66\u4e60\u800c\u6765\u7684\u65b9\u6cd5<\/span>\uff08\u89c1\u88683\uff09<\/span>\u3002\u5728\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u9886\u57df\uff0c\u7ed3\u6784\u6027\u56e0\u679c\u6a21\u578b\u7684\u6846\u67b6\u6709\u5f88\u5927\u7684\u5e94\u7528\u6f5c\u529b\uff0c\u5373\u4fbf\u5f88\u5927\u7a0b\u5ea6\u4e0a\u8fd8\u6ca1\u6709\u88ab\u5f00\u53d1\u51fa\u6765\u3002<\/span><\/section>\n

<\/span><\/section>\n
\u88683 \u7efc\u8ff0\u4e0e\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u4e2d\u6700\u76f8\u5173\u7684\u673a\u5668\u5b66\u4e60\u65b9\u6cd5\u975e\u6280\u672f\u63cf\u8ff0<\/span><\/section>\n\n\n\n\n\n\n\n
\n
\u7528\u4e8e\u56e0\u679c\u63a8\u65ad\u7684\u673a\u5668\u5b66\u4e60<\/span><\/strong><\/section>\n<\/td>\n<\/tr>\n
\n
\u53cd\u4e8b\u5b9e\u9884\u6d4b<\/span><\/strong><\/section>\n<\/td>\n<\/tr>\n
\n
\u8bbe\u7f6e\u6570\u636e\u4ee5\u6ee1\u8db3\u56e0\u679c\u63a8\u65ad\u7684\u5047\u8bbe\uff0c\u5e76\u4f7f\u7528\u9884\u6d4b\u6027ML\uff08\u88682\uff09\u8fdb\u884c\u6548\u5e94\u4f30\u8ba1\u3002\u4ee5\u4e0b\u662f\u51e0\u4e2a\u63d0\u4f9b\u4e86\u652f\u6301\u56e0\u679c\u6548\u5e94\u4f30\u8ba1\u547d\u4ee4\u7684\u8f6f\u4ef6\u5305\uff1a<\/span><\/section>\n
1\uff09BART\uff1a\u8be6\u60c5\u89c1\u524d\u6587\u63cf\u8ff0<\/span><\/section>\n
2\uff09\u76ee\u6807\u6700\u5927\u4f3c\u7136\u4f30\u8ba1\u6cd5\uff08TMLE\uff09\uff1a\u5728\u4e8c\u5143\u70b9\u6cbb\u7597\u548c\u4e8c\u5143\u7ed3\u679c\u7684\u57fa\u672c\u8bbe\u7f6e\u4e2d\uff0c\u8be5\u65b9\u6cd5(i)\u4f30\u8ba1\u63a5\u53d7\u6cbb\u7597\u548c\u672a\u63a5\u53d7\u6cbb\u7597\u7684\u89c2\u5bdf\u503c\uff08\u5982\u4e2a\u4eba\uff09\u7684\u9884\u671f\u7ed3\u679c\u3001(ii)\u4f30\u8ba1\u6240\u6709\u89c2\u5bdf\u503c\u63a5\u53d7\u6cbb\u7597\u7684\u6982\u7387\u3001(iii)\u4f30\u8ba1\u4e00\u4e2a\u6ce2\u52a8\u53c2\u6570\u4ee5\u544a\u77e5\u66f4\u65b0\u521d\u59cb\u7ed3\u679c\u4f30\u8ba1\u7684\u7a0b\u5ea6\u3001(iv)\u66f4\u65b0\u521d\u59cb\u7ed3\u679c\u4f30\u8ba1\u3001(v)\u8ba1\u7b97\u5e73\u5747\u6cbb\u7597\u6548\u679c\u548c\u6807\u51c6\u8bef\u5dee\uff08SE\uff09\u7528\u4e8e\u7edf\u8ba1\u63a8\u7406\u3002<\/span><\/section>\n
3\uff09\u56e0\u679c\u68ee\u6797\uff1a\u4e00\u79cd\u57fa\u4e8e\u68ee\u6797\u7edf\u8ba1\u63a8\u7406\u7684\u6cbb\u7597\u6548\u679c\u4f30\u8ba1\u65b9\u6cd5\uff08105\uff09\u3002\u5173\u4e8e\u968f\u673a\u68ee\u6797\u7684\u8be6\u7ec6\u89e3\u91ca\uff0c\u89c1\u88682\u3002<\/span><\/section>\n
4\uff09\u53cc\u91cd\u673a\u5668\u5b66\u4e60\uff08DML\uff09\uff1a\u5728\u4e8c\u5143\u5904\u7406\u7684\u914d\u7f6e\u4e2d\uff0c\u5bf9\u4e8e\u53cc\u91cd\u4ea4\u53c9\u62df\u5408\u7684\u60c5\u51b5\uff0c\u8981\u8fdb\u884c\u4ee5\u4e0b\u6b65\u9aa4\u3002(i) \u8be5\u65b9\u6cd5\u5c06\u6570\u636e\u968f\u673a\u5206\u6210\u4e24\u7ec4\uff1b(ii) \u5bf9\u4e8e\u7b2c\u4e00\u7ec4\uff0c\u4f7f\u7528ML\u5728\u534f\u53d8\u91cf\u7684\u57fa\u7840\u4e0a\u9884\u6d4b\u7ed3\u679c\uff1b(iii) \u5bf9\u4e8e\u7b2c\u4e00\u7ec4\uff0c\u4f7f\u7528ML\u5728\u534f\u53d8\u91cf\u7684\u57fa\u7840\u4e0a\u9884\u6d4b\u6cbb\u7597\uff1b(iv) \u5c06(ii)\u5f97\u5230\u7684\u7ed3\u679c\u6b8b\u5dee\u56de\u5f52\u5230\u6cbb\u7597\u6b8b\u5dee(iii)\uff0c\u5f97\u5230\u4e00\u4e2a\u6a21\u578b\u3002(v) \u5bf9\u4e8e\u7b2c\u4e8c\u7ec4\uff0c\u4f7f\u7528\u4ece(iv)\u4e2d\u5f97\u5230\u7684\u6a21\u578b\u6765\u5f97\u5230\u6cbb\u7597\u6548\u679c\u7684\u4f30\u8ba1\u503c\uff1b(vi) \u5bf9\u4e8e\u7b2c\u4e8c\u7ec4\u91cd\u590d\u540c\u6837\u7684\u7a0b\u5e8f\uff1b(vii) \u5bf9\u4ece\u4e24\u7ec4\u5f97\u5230\u7684\u6cbb\u7597\u4f30\u8ba1\u503c\u8fdb\u884c\u5e73\u5747\uff0c\u8fdb\u884c\u6700\u7ec8\u4f30\u8ba1(106)\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
\u56e0\u679c\u7ed3\u6784\u5b66\u4e60<\/span><\/strong><\/section>\n<\/td>\n<\/tr>\n
\n
1\uff09\u57fa\u4e8e\u7ea6\u675f\u7684\u7b97\u6cd5\uff1a\u9996\u5148\u5b66\u4e60\u4e00\u4e2a\u65e0\u5411\u56fe\uff0c\u7136\u540e\u786e\u5b9a\u65b9\u5411\uff0c\u5305\u62ec\u4ee5\u4e0b\u51e0\u79cd\u4f8b\u5b50\uff1a(i) \u7a33\u5b9a\u7684PC\u7b97\u6cd5\uff08142\uff09\u901a\u8fc7\u8fdb\u884c\u5927\u91cf\u7684\u6761\u4ef6\u72ec\u7acb\u6027\u6d4b\u8bd5\u6765\u5b66\u4e60DAG\u7ed3\u6784\uff0c\u5e76\u89e3\u51b3\u4e86\u7ecf\u5178PC\u7b97\u6cd5\uff08143\uff09\u4e2d\u7684\u987a\u5e8f\u4f9d\u8d56\u95ee\u9898\uff1b(ii) \u5feb\u901f\u56e0\u679c\u63a8\u7406\uff08144\uff09\u662fPC\u7b97\u6cd5\u7684\u6269\u5c55\uff0c\u5f53\u56e0\u679c\u5145\u5206\u6027\u7684\u5047\u8bbe\u53ef\u80fd\u4e0d\u88ab\u6ee1\u8db3\u65f6\uff0c\u53ef\u8bf4\u660e\u53ef\u80fd\u7684\u6f5c\u5728\u53d8\u91cf\u3002(iii) MMPC\u662f\u4e00\u79cd\u4e24\u9636\u6bb5\u7684\u7b97\u6cd5\uff0c\u7528\u4e8e\u5b66\u4e60\u7f51\u7edc\u4e2d\u4efb\u4f55\u53d8\u91cf\u7684\u76f4\u63a5\u539f\u56e0\u548c\u5f71\u54cd(145)\uff1b\u4ee5\u53ca(iv) IAMB (146)\u4f7f\u7528Markov blanket\u7684\u6982\u5ff5\u6765\u51cf\u5c11\u6761\u4ef6\u72ec\u7acb\u6027\u6d4b\u8bd5\u7684\u6570\u91cf\uff08\u5bf9\u4e8e\u4e00\u4e2a\u7279\u5b9a\u7684\u53d8\u91cf\uff0cMarkov blanket\u662f\u6700\u5c0f\u7684\u6761\u4ef6\u96c6\uff0c\u786e\u4fdd\u7279\u5b9a\u8282\u70b9\u6709\u6761\u4ef6\u5730\u72ec\u7acb\u4e8e\u56fe\u4e2d\u7684\u6240\u6709\u5176\u4ed6\u8282\u70b9\uff09\u3002<\/span><\/section>\n
2\uff09\u57fa\u4e8e\u5206\u6570\u7684\u7b97\u6cd5\uff1a\u5728DAG\u4e0a\u641c\u7d22\uff0c\u5e76\u6839\u636e\u6307\u5b9a\u7684\u76ee\u6807\u51fd\u6570\u5bf9\u6bcf\u4e2aDAG\u8fdb\u884c\u8bc4\u5206\uff0c\u5305\u62ec\u4ee5\u4e0b\u51e0\u79cd\u4f8b\u5b50\uff1a(i)   \u722c\u5761\uff1a\u4ece\u4e00\u4e2a\u7a7a\u56fe\u5f00\u59cb\uff0c\u8fed\u4ee3\u5730\u8fdb\u884c\u5c40\u90e8\u7684\u589e\u52a0\u3001\u5220\u9664\u6216\u98a0\u5012\u8fb9\u7f18\uff0c\u4ee5\u63d0\u9ad8\u56fe\u7684\u5f97\u5206\uff08147\uff09\uff1b(ii) Tabu\u641c\u7d22\u662f\u722c\u5761\u7684\u6269\u5c55\uff0c\u8bd5\u56fe\u5e2e\u52a9\u7b97\u6cd5\u6446\u8131\u5c40\u90e8\u6a21\u5f0f\uff08147\uff09\uff1b(iii) \u5feb\u901f\u8d2a\u5a6a\u7b49\u4ef7\u641c\u7d22\uff1a\u589e\u52a0\u548c\u5220\u9664\u591a\u9879\u5f0f\u7684\u8fb9\u7f18\uff0c\u5728DAG\u7684\u7b49\u4ef7\u7c7b\u7a7a\u95f4\u4e2d\u641c\u7d22\uff08148\uff09\u3002<\/span><\/section>\n
3\uff09\u6df7\u5408\u65b9\u6cd5\uff1a\u7ed3\u5408\u57fa\u4e8e\u7ea6\u675f\u7684\u548c\u57fa\u4e8e\u5206\u6570\u7684\u65b9\u6cd5\u3002\u4f8b\u5b50\uff1aMMHC (149)\u4f7f\u7528\u57fa\u4e8e\u7ea6\u675f\u7684MMPC\u6765\u5efa\u7acbDAG\u7684\u9aa8\u67b6\uff0c\u7136\u540e\u4f7f\u7528\u57fa\u4e8e\u5206\u6570\u7684\u722c\u5761\u6765\u786e\u5b9a\u65b9\u5411\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

\n

<\/section>\n
6.1 \u5e94\u75281\uff1a\u53cd\u4e8b\u5b9e\u7684\u63a8\u6d4b<\/span><\/strong><\/span><\/section>\n

<\/span><\/section>\n<\/h3>\n
\u6839\u636eHern\u00e1n\u7b49\u4eba\uff086\uff09\u7684\u8bf4\u6cd5\uff0c\u201c\u53cd\u4e8b\u5b9e\u9884\u6d4b\u662f\u5229\u7528\u6570\u636e\u6765\u9884\u6d4b\u4e16\u754c\u7684\u67d0\u4e9b\u7279\u5f81\uff0c\u5c31\u50cf\u4e16\u754c\u5df2\u7ecf\u4e0d\u540c\u4e00\u6837\u201d\u3002\u5982\u679c\u5bfb\u6c42\u9884\u6d4b\u6216\u53cd\u4e8b\u5b9e\u9884\u6d4b\uff0c\u5224\u65ad\u7b54\u6848\u662f\u5426\u6709\u7528\u5f97\u5230\u4e00\u4e2a\u6807\u51c6\u662f\u95ee\u76ee\u6807\u662f\u201c\u89e3\u91ca\u8fd8\u662f\u9884\u6d4b\uff1f\u201d(100)\u3002\u5c3d\u7ba1\u5728\u67d0\u4e9b\u60c5\u51b5\u4e0b\uff0c\u53ef\u80fd\u80fd\u591f\u4f30\u8ba1\u56e0\u679c\u6548\u5e94\u7684\u5927\u5c0f\uff0c\u4f46\u4ecd\u65e0\u6cd5\u89e3\u91ca<\/span>\uff08\u4f8b\u5982\uff0c\u5728\u4e00\u4e2a\u8bd5\u9a8c\u4e2d\uff09<\/span>(6)\u3002\u6f5c\u5728\u7ed3\u679c\u6846\u67b6\u4e2d\u7684\u56e0\u679c\u95ee\u9898<\/span>\uff08\u7528\u4e8e\u53cd\u4e8b\u5b9e\u9884\u6d4b\uff09<\/span>\u53ef\u4ee5\u7528\u4f20\u7edf\u7684\u65b9\u6cd5\u6765\u56de\u7b54\uff0c\u5982\u56de\u5f52\u4ee5\u53ca\u66f4\u5148\u8fdb\u7684\u5982\u8fb9\u9645\u7ed3\u6784\u6a21\u578b\uff086\uff09\u7684\u65b9\u6cd5\u3002\u7136\u800c\uff0c\u5728\u8fc7\u53bb\u7684\u51e0\u5e74\u91cc\uff0c\u7528\u4e8e\u56e0\u679c\u63a8\u65ad\u7684ML\u5df2\u7ecf\u6709\u4e86\u5f88\u5927\u7684\u53d1\u5c55\uff0c\u5982\u679c\u80fd\u591f\u5d4c\u5165\u5230\u56e0\u679c\u63a8\u65ad\u548c\u7edf\u8ba1\u63a8\u65ad\u7684\u6846\u67b6\u4e2d\uff0c\u5c06\u7279\u522b\u6709\u5e2e\u52a9\uff08101\uff09\u3002<\/span><\/section>\n

<\/span><\/section>\n

\n
6.1.1 \u5e38\u89c1\u7684\u673a\u5668\u5b66\u4e60\u65b9\u6cd5<\/span><\/strong><\/section>\n

<\/span><\/section>\n<\/h4>\n
\u5bf9\u4e8e\u53cd\u4e8b\u5b9e\u9884\u6d4b\uff0cML\u53ef\u4ee5\u7528\u6765\u5206\u6790\u5927\u91cf\u7684\u6570\u636e\u96c6\u3002\u901a\u8fc7\u8bbe\u7f6e\u6570\u636e\u7684\u65b9\u5f0f\u6765\u6ee1\u8db3\u56e0\u679c\u5047\u8bbe\uff0c\u7136\u540e\u4f7f\u7528\u9884\u6d4b\u6027ML\u65b9\u6cd5\u6765\u56de\u7b54\u56e0\u679c\u95ee\u9898\uff08102\uff09\u3002\u5982\u679c\u6570\u636e\u8bbe\u7f6e\u5f97\u5f53\uff0c\u5176\u4ed6ML\u65b9\u6cd5\u4e5f\u5f88\u9002\u5408\u89e3\u51b3\u56e0\u679c\u95ee\u9898\uff0c\u5982BART\uff08103\uff09\u3001TMLE\uff08104\uff09\u548c\u968f\u673a\u68ee\u6797\uff08105\uff09\u3002\u53e6\u4e00\u79cd\u56de\u7b54\u56e0\u679c\u95ee\u9898\u7684ML\u65b9\u6cd5\u662f\u53cc\u91cd\u673a\u5668\u5b66\u4e60<\/span>\uff08DML\uff09<\/span>\uff0c\u5b83\u5728\u9ad8\u7ef4\u6570\u636e\u4e2d\u5b66\u4e60\u5e73\u5747\u6cbb\u7597\u6548\u679c\u548c\u5bf9\u88ab\u6cbb\u7597\u8005\u7684\u5e73\u5747\u6cbb\u7597\u6548\u679c\uff08106\uff09\u3002DML\u80fd\u591f\u5229\u7528ML\u65b9\u6cd5\u7684\u51c6\u786e\u9884\u6d4b\uff0c\u540c\u65f6\u901a\u8fc7\u6837\u672c\u5206\u5272\u548c\u4ea4\u53c9\u62df\u5408\u63d0\u4f9b\u65e0\u504f\u7684\u3001\u4e00\u81f4\u7684\u4f30\u8ba1\uff0c\u5e76\u63d0\u4f9b\u6709\u6548\u7684\u7edf\u8ba1\u63a8\u65ad\u3002\u5f3a\u5316\u5b66\u4e60\u4e5f\u53ef\u4ee5\u88ab\u6982\u5ff5\u5316\u4e3a\u4e00\u79cd\u65b9\u6cd5\uff0c\u5728\u6709\u53ef\u80fd\u4ea7\u751f\u6570\u636e<\/span>\uff08A\/B\u6d4b\u8bd5\uff09<\/span>\u7684\u7814\u7a76\u80cc\u666f\u4e0b\u63a5\u8fd1\u56e0\u679c\u63a8\u65ad\u3002\u88683\u7ed9\u51fa\u4e86\u4f7f\u7528\u6f5c\u5728\u7ed3\u679c\u6846\u67b6\u8fdb\u884c\u56e0\u679c\u63a8\u65ad\u7684\u5e38\u89c1ML\u65b9\u6cd5\u7684\u6982\u8ff0\u3002<\/span><\/section>\n

<\/span><\/section>\n
\u5728\u67d0\u4e9b\u60c5\u51b5\u4e0b\uff0c\u5df2\u77e5\u53d8\u91cf\u7684\u56e0\u679c\u7ed3\u6784<\/span>\uff08\u90e8\u5206\uff09<\/span>\uff0c\u4f8b\u5982\u5982\u679c\u53d8\u91cf\u65e2\u76f8\u4e92\u8054\u7cfb\u53c8\u6709\u65f6\u95f4\u987a\u5e8f\uff0c\u6216\u8005\u9075\u5faa\u53e6\u4e00\u79cd\u56e0\u679c\u903b\u8f91<\/span>\uff08\u4f8b\u5982\uff0c\u7814\u7a76\u8005\u5bf9\u81ea\u53d8\u91cf\u6216\u53e6\u4e00\u4e2a\u5916\u5728\u539f\u56e0\u7684\u64cd\u7eb5\uff09<\/span>\u3002\u5728\u6570\u636e\u79d1\u5b66\u9886\u57df\uff0c\u4eba\u4eec\u4f1a\u63d0\u53ca\u7ed3\u6784\u5316\u7684\u9ad8\u7ef4\u8f93\u5165\u3002\u5728\u8fd9\u4e9b\u60c5\u51b5\u4e0b\uff0c\u53ef\u4ee5\u4f7f\u7528ML\u4e2d\u7684\u56fe\u6765\u7eb3\u5165\u8fd9\u79cd\u56e0\u679c\u77e5\u8bc6\uff0c\u5e76\u6269\u5c55LASSO\u6216\u795e\u7ecf\u7f51\u7edc\u7b49\u9884\u6d4b\u7b97\u6cd5\uff0c\u4f8b\u5982\u5229\u7528LASSO\u6216\u5377\u79ef\u795e\u7ecf\u7f51\u7edc\u6765\u53cd\u6620\u56e0\u679c\u7ed3\u6784\u3002<\/span><\/section>\n

<\/span><\/section>\n
BART\u7684\u8be6\u7ec6\u8bf4\u660e\uff1a\u672c\u6587\u4ecb\u7ecd\u7684BART\u662f\u89e3\u51b3\u53cd\u4e8b\u5b9e\u9884\u6d4b\u7684\u4e00\u79cd\u6709\u8da3\u7684\u65b9\u6cd5\uff0c\u56e0\u4e3a\u5b83\u7ed3\u5408\u4e86\u5178\u578b\u7684\u56e0\u679c\u63a8\u65ad\u65b9\u6cd5\uff0c\u4f8b\u5982\u503e\u5411\u6027\u8bc4\u5206\u52a0\u6743\u6765\u5e73\u8861\u6cbb\u7597\u5206\u914d\u7684\u6982\u7387\u4e0e\u6df7\u6742\u56e0\u7d20\u6765\u8ba1\u7b97\u53cd\u4e8b\u5b9e\uff0c\u5e76\u7528\u6765\u4f30\u8ba1\u6cbb\u7597\u7684\u6548\u679c\uff0c\u6240\u4ee5\u8bb8\u591a\u5fc5\u8981\u7684\u7814\u7a76\u8005\u51b3\u5b9a\u662f\u660e\u786e\u7684\u3002\u66f4\u8fdb\u4e00\u6b65\uff0cBART\u662f\u6811\u4e4b\u548c\u7684\u6a21\u578b\uff0c\u9884\u6d4b\u7ed3\u679c\u662f\u5355\u4e2a\u56de\u5f52\u6811\u62df\u5408\u7684\u96c6\u5408\u548c\u4e00\u4e2a\u52a0\u6027\u9ad8\u65af\u8bef\u5dee\u9879\u7684\u603b\u548c\u3002\u6bcf\u4e2a\u56de\u5f52\u6811\u8fed\u4ee3\u5730\u5e94\u7528\u5206\u5272\u89c4\u5219\uff0c\u5c06\u6570\u636e\u5206\u5272\u6210\u4e0d\u91cd\u53e0\u7684\u5b50\u96c6\uff0c\u76ee\u7684\u662f\u4f7f\u6bcf\u4e2a\u5b50\u96c6\u7684\u65b9\u5dee\u6700\u5c0f\u5316\uff0855, 103\uff09\u3002\u7531\u4e8e\u5355\u4e2a\u6811\u8fc7\u5ea6\u5f3a\u8c03\u76f8\u4e92\u4f5c\u7528\uff0c\u96be\u4ee5\u8bc6\u522b\u771f\u6b63\u7684\u7ebf\u6027\u5173\u7cfb\uff0c\u56e0\u6b64\u540e\u7eed\u7684\u6811\u662f\u6839\u636e\u8bc6\u522b\u5b50\u96c6\u7684\u6b8b\u5dee\u9884\u6d4b\u503c\u6765\u62df\u5408\u7684\uff08103\uff09\u3002\u4e3a\u4e86\u907f\u514d\u8fc7\u5ea6\u62df\u5408\uff0cBART\u5f15\u5165\u4e86\u6811\u7684\u5927\u5c0f<\/span>\uff08\u5373\u5b50\u96c6\/\u7ec8\u7aef\u8282\u70b9\u7684\u6570\u91cf\uff09<\/span>\u548c\u6536\u7f29<\/span>\uff08\u5373\u5728\u5b50\u96c6\u4e2d\u5e73\u6574\u7684\u56e0\u7d20\uff09<\/span>\u7684\u6b63\u5219\u5316\u5148\u9a8c\u3002\u7136\u800c\uff0c\u6811\u7684\u6570\u91cf\u4ecd\u7136\u662fBART\u6a21\u578b\u7684\u4e00\u4e2a\u8c03\u6574\u53c2\u6570\uff08107\uff09\u3002\u5bf9\u5e95\u5c42\u8d1d\u53f6\u65af\u53cd\u62df\u5408\u7b97\u6cd5\u7684\u66f4\u8be6\u7ec6\u63cf\u8ff0\u53ef\u4ee5\u5728\u5176\u4ed6\u5730\u65b9\u627e\u5230(103)\u3002BART\u7684\u5b9e\u65bd\u5728\u56de\u5f52\u548c\u5206\u7c7b\u8bbe\u7f6e\u4e2d\u90fd\u5b58\u5728\u3002\u4e0e\u5e38\u89c1\u7684\u57fa\u4e8e\u6811\u7684ML\u65b9\u6cd5<\/span>\uff08\u5982\u968f\u673a\u68ee\u6797\u6216\u63d0\u5347\uff09<\/span>\u4e0d\u540c\uff0c\u6b63\u5219\u5316\u4f18\u5148\u6743\u4f20\u8fbe\u4e86\u7075\u6d3b\u7684\u8d8b\u52bf<\/span>\uff08\u5982\u5bf9\u5c0f\u6811\uff09<\/span>\uff0c\u800c\u4e0d\u662f\u7531\u8ba1\u7b97\u91cf\u5927\u7684\u7f51\u683c\u641c\u7d22\u786e\u5b9a\u7684\u56fa\u5b9a\u53c2\u6570\u3002\u4f18\u5148\u6743\u8fdb\u4e00\u6b65\u9002\u7528\u4e8e\u9ad8\u7ef4\u6570\u636e\u548c\u5e73\u6ed1\u56de\u5f52\u51fd\u6570\uff0855\uff09\u3002\u5b9e\u9a8c\u7ed3\u679c\u8868\u660e\uff0c\u5176\u6027\u80fd\u53ef\u4ee5\u4e0e\u666e\u901a\u7684\u65b9\u6cd5\u76f8\u5ab2\u7f8e\u6216\u8d85\u8fc7\uff0c\u5982\u63d0\u5347\u3001\u795e\u7ecf\u7f51\u7edc\u3001\u6216\u968f\u673a\u68ee\u6797\u3002\u7136\u800c\uff0c\u7279\u522b\u662f\u5bf9\u4e8e\u4e8c\u5143\u7ed3\u679c\uff0c\u4ea4\u53c9\u9a8c\u8bc1BART\u6a21\u578b\u4ee5\u9009\u62e9\u6b63\u5219\u5316\u5148\u9a8c\u662f\u6709\u5229\u7684\uff0855\uff0c103\uff09\u3002\u9664\u4e86\u8ba1\u7b97\u65b9\u9762\u7684\u597d\u5904\u5916\uff0cBART \u8fd8\u9002\u7528\u4e8e\u5404\u79cd\u7814\u7a76\u91cd\u70b9\u548c\u7ed3\u679c<\/span>\uff08\u5982\u751f\u5b58\u548c\u591a\u53c9\u903b\u8f91\u56de\u5f52\uff09<\/span>\uff0c\u7279\u522b\u9002\u5408\u4e8e\u56e0\u679c\u63a8\u65ad\u4efb\u52a1\uff0c\u56e0\u4e3a\u590d\u6742\u53cd\u5e94\u9762\u7684\u5efa\u6a21\u548c\u63a7\u5236\u6df7\u6742\u4e0d\u4f9d\u8d56\u4e8e\u53c2\u6570\u5316\u7684\u5047\u8bbe\uff0855, 107\uff09\u3002\u7531\u6b64\u4ea7\u751f\u7684\u540e\u9a8c\u5206\u5e03\u5141\u8bb8\u4f30\u8ba1\u4e2a\u4eba\u5e73\u5747\u6cbb\u7597\u6216\u5f02\u8d28\u6027\u56e0\u679c\u6548\u5e94\u3002\u6b64\u5916\uff0c\u57fa\u672c\u7684\u4f3c\u7136\u6846\u67b6\u63d0\u4f9b\u4e86\u5173\u4e8e\u7ed3\u679c\u7684\u6982\u7387\u58f0\u660e\uff0c\u5305\u62ec\u53ef\u4fe1\u533a\u95f4\uff0c\u800c\u8bc6\u522b\u548c\u91cf\u5316\u5355\u4e2a\u53d8\u91cf\u5bf9\u7ed3\u679c\u7684\u5f71\u54cd\u5219\u66f4\u4e3a\u590d\u6742\uff08107\uff09\u3002\u6700\u8fd1\u5bf9BART\u7684\u6539\u7f16\u548c\u5b9e\u65bd\u8fdb\u4e00\u6b65\u5141\u8bb8\u5bf9\u6cbb\u7597\u7684\u6982\u7387\u8fdb\u884c\u5efa\u6a21\u5e76\u5305\u62ec\u8bc4\u5206\uff0c\u4ee5\u53ca\u5728\u5b58\u5728\u672a\u89c2\u5bdf\u5230\u7684\u6df7\u6742\u56e0\u7d20\u65f6\u6a21\u62df\u6cbb\u7597\u6548\u679c\u3002\u9664\u6b64\u4ee5\u5916\uff0c\u8fd8\u6709\u4e00\u4e9b\u7a0b\u5e8f\u7f3a\u5c11\u5171\u8bc6\u7684\u7a0b\u5e8f\u53ef\u4ee5\u63a7\u5236(55)\u3002<\/span><\/section>\n

<\/span><\/section>\n
\u793e\u4f1a\u79d1\u5b66\u548c\u5065\u5eb7\u79d1\u5b66\u4e2d\u5e38\u89c1\u7684\u56e0\u679c\u63a8\u65ad\u7684\u7814\u7a76\u76ee\u7684\uff0c\u901a\u5e38\u7528\u6f5c\u5728\u7684\u7ed3\u679c\u6846\u67b6\u6765\u89e3\u51b3\uff0c\u5305\u62ec(i)\u8bc4\u4f30\u4e0d\u5229\u7684\u793e\u4f1a\u6216\u5065\u5eb7\u7ed3\u679c\u7684\u6f5c\u5728\u539f\u56e0\u3001(ii)\u8bc4\u4f30\u6bd4\u8f83\u6cbb\u7597\u6548\u679c\u3001(iii)\u8bc6\u522b\u5f02\u8d28\u6027\u6cbb\u7597\u6548\u679c\u4ee5\u53ca(iv)\u8bc4\u4f30\u548c\u53ef\u80fd\u6d88\u9664\u7edf\u8ba1\u5206\u6790\u4e2d\u7684\u504f\u89c1\u3002<\/span><\/section>\n

<\/span><\/section>\n

\n
6.1.2 \u5bf9\u793e\u4f1a\u548c\u5065\u5eb7\u7684\u6f5c\u5728\u56e0\u7d20\u8bc4\u4f30<\/span><\/strong><\/section>\n

<\/span><\/section>\n<\/h4>\n
\u4e3a\u4e86\u8bc4\u4f30\u6f5c\u5728\u56e0\u7d20\u7684\u6548\u679c\uff0c\u53ef\u4ee5\u5728\u63a7\u5236\u6df7\u6742\u56e0\u7d20\u7684\u60c5\u51b5\u4e0b\u8bbe\u8ba1\u7814\u7a76\uff0c\u4ee5\u8bc4\u4f30\u9884\u6d4b\u8005\u5bf9\u56e0\u679c\u7684\u5f71\u54cd\u3002\u4e00\u9879\u7814\u7a76\u6d4b\u8bd5\u4e86\u51c6\u5988\u5988\u8425\u517b\u4e2d\u7684\u6c34\u679c\/\u852c\u83dc\u5bc6\u5ea6\u5bf9\u4e0d\u826f\u598a\u5a20\u548c\u5206\u5a29\u7ed3\u679c\u7684\u5f71\u54cd\uff0c\u7ed3\u679c\u663e\u793a\uff0cTMLE\u901a\u8fc7\u53d1\u73b0\u5c0f\u7684\u5f71\u54cd\u548c\u7ed9\u51fa\u66f4\u7cbe\u786e\u7684\u4f30\u8ba1\uff0c\u80dc\u8fc7\u4e86\u4f20\u7edf\u7684\u5efa\u6a21\uff08108\uff09\u3002\u5728\u8fd9\u4e2a\u95ee\u9898\u4e0a\u66f4\u8fdb\u4e00\u6b65\uff0c\u540e\u7eed\u7814\u7a76\u4ee5\u53cc\u91cd\u7a33\u5065\u7684ML\u663e\u793a\uff0c\u6c34\u679c\u548c\u852c\u83dc\u5bc6\u5ea6\u5bf9\u5b50\u75eb\u524d\u671f\u98ce\u9669\u7684\u4fdd\u62a4\u4f5c\u7528\u88ab\u5373\u5c06\u6210\u4e3a\u6bcd\u4eb2\u7684BMI\u6240\u6539\u53d8\uff0c\u4fdd\u62a4\u4f5c\u7528\u968f\u7740BMI\u572820-30\u5206\u4e4b\u95f4\u7684\u589e\u52a0\u800c\u52a0\u5f3a\uff0c\u800c\u5bf9\u4e8eBMI\u572830\u5206\u4ee5\u4e0a\u7684\u5373\u5c06\u6210\u4e3a\u6bcd\u4eb2\u7684\u4eba\u6765\u8bf4\uff0c\u6548\u679c\u8d8b\u4e8e\u5e73\u7a33\uff08109\uff09\u3002\u5e94\u7528\u56e0\u679c\u89c2\u70b9\u751a\u81f3\u6709\u52a9\u4e8e\u89e3\u51b3\u5371\u91cd\u75c5\u4eba\u7684\u80a5\u80d6\u6096\u8bba\uff08110\uff09\u3002\u4f5c\u4e3a\u793e\u4f1a\u6d41\u52a8\u7814\u7a76\u7684\u4e00\u4e2a\u4f8b\u5b50\uff0c\u7528\u56de\u5f52\u6811\u7684\u7eb5\u5411\u89c6\u89d2\u6bd4\u8f83\u4e86\u7236\u6bcd\u548c\u4e2a\u4eba\u7684\u793e\u4f1a\u7ecf\u6d4e\u6536\u5165\u51b3\u5b9a\u56e0\u7d20(111)\u3002<\/span><\/section>\n

<\/span><\/section>\n

\n
6.1.3 \u8bc4\u4f30\u6bd4\u8f83\u6cbb\u7597\u7684\u6709\u6548\u6027<\/span><\/strong><\/section>\n

<\/span><\/section>\n<\/h4>\n
\u672c\u8282\u4e2d\u7814\u7a76\u4eba\u5458\u53ef\u8bc4\u4f30<\/span>\uff08\u4f8b\u5982\u5e72\u9884\u4e0e\u63a7\u5236\u6216\u7167\u6599\uff09<\/span>\u5728\u6539\u53d8\u5065\u5eb7\u7ed3\u679c\u65b9\u9762\u6700\u6709\u6548\u7684\u51e0\u79cd\u6cbb\u7597\u65b9\u6cd5\u3002\u5173\u4e8e\u536b\u751f\u670d\u52a1\u7814\u7a76\u4e2d\u66f4\u591a\u4e34\u5e8a\u5e94\u7528\u7684\u9009\u62e9\u6027\u6982\u8ff0\u89c1(112)\u3002\u7531\u4e8e\u8bc4\u4f30\u201c\u6cbb\u7597\u201d\u5728\u653f\u7b56\u8bc4\u4f30\u4e2d\u4e5f\u6709\u76f8\u4f3c\u4e4b\u5904\uff0c\u6211\u4eec\u5728(113)\u4e2d\u56de\u987e\u4e86\u7ecf\u6d4e\u5b66\u4e2d\u57fa\u4e8eML\u4f30\u8ba1\u5e73\u5747\u6cbb\u7597\u6548\u5e94\u7684\u53d1\u5c55\u3002\u4e00\u9879\u7814\u7a76\u6d4b\u8bd5\u4e86\u8ba4\u77e5\u884c\u4e3a\u7597\u6cd5\u7684\u76ee\u6807\u5904\u65b9\uff0c\u800c\u4ee5\u4f30\u8ba1\u6cbb\u7597\u7ed3\u679c\u4e3a\u4e2d\u5fc3\u7684\u4eba\u88ab\u5206\u914d\u5230\u53e6\u4e00\u79cd\u6cbb\u7597(=\u53cd\u4e8b\u5b9e\u7ed3\u679c) (114)\u3002<\/span><\/section>\n

<\/span><\/section>\n

\n
6.1.4 \u8bc6\u522b\u5f02\u8d28\u6027\u6cbb\u7597\u6548\u679c<\/span><\/strong><\/section>\n

<\/span><\/section>\n<\/h4>\n
\u5728\u968f\u673a\u5904\u7406\u5206\u914d\u7684\u60c5\u51b5\u4e0b\uff0c\u6216\u8005\u5728\u8bc4\u4f30\u65b0\u5f15\u5165\u6216\u6539\u53d8\u7684\u653f\u7b56\u6548\u679c\u7684\u65f6\u95f4\u7814\u7a76\u8bbe\u8ba1\u4e2d\uff0c\u53cd\u4e8b\u5b9e\u9884\u6d4b\u4f30\u8ba1\u5e73\u5747\u5904\u7406\u6548\u5e94\u662f\u76f4\u63a5\u7684\u3002\u6b64\u5916\uff0c\u7814\u7a76\u8005\u53ef\u80fd\u5e0c\u671b\u8bc6\u522b\u548c\u63cf\u8ff0\u5bf9\u6cbb\u7597\u53cd\u5e94\u4e0d\u540c\u7684\u4e9a\u7ec4\uff0c\u5373\u63a2\u8ba8\u5f02\u8d28\u6027\u7684\u6cbb\u7597\u6548\u679c\u3002\u5f02\u8d28\u6027\u5904\u7406\u6548\u5e94\u7684\u5206\u6790\u53ef\u4ee5\u56de\u7b54\u516c\u5171\u536b\u751f\u7814\u7a76\u4e2d\u5c24\u4e3a\u7a81\u51fa\u7684\u95ee\u9898\uff1a\u4ec0\u4e48\u5bf9\u8c01\u6700\u6709\u6548\uff0c\u4ec0\u4e48\u65f6\u5019\u6700\u6709\u6548\uff1f\u6280\u672f\u8be6\u7ec6\u8bf4\u660e\u8bf7\u89c1\u4f30\u8ba1\u5904\u7406\u6548\u5e94\u5f02\u8d28\u6027\u7684\u6587\u732e(115)\u3002\u968f\u673a\u68ee\u6797\u5df2\u88ab\u5f00\u53d1\u7528\u4e8e\u68c0\u6d4b\u5f02\u8d28\u6027\u5904\u7406\u6548\u5e94(105)\u3002DML\u4e5f\u88ab\u7528\u6765\u4f30\u8ba1\u5f02\u8d28\u6027\u5904\u7406\u6548\u5e94(106)\u3002<\/span><\/section>\n

<\/span><\/section>\n
\u867d\u7136\u4e34\u5e8a\u8bd5\u9a8c\u7684\u518d\u5206\u6790\u662f\u4e3a\u4e86\u66f4\u597d\u5730\u7406\u89e3\u5bf9\u5e72\u9884\u6216\u836f\u7269\u6cbb\u7597\u7684\u53cd\u5e94\u4e2d\u53ef\u80fd\u5b58\u5728\u7684\u5f02\u8d28\u6027\uff0c\u4f46\u4ecd\u7136\u9700\u8981\u8c28\u614e\uff1a\u91cd\u65b0\u5206\u6790\u5931\u8d25\u7684\u8bd5\u9a8c\u7684\u5f02\u8d28\u6027\u6cbb\u7597\u6548\u5e94\u53ef\u4ee5\u88ab\u8ba4\u4e3a\u662f\u6709\u95ee\u9898\u7684\u548cP\u503c\u64cd\u7eb5\u7684\u5f62\u5f0f\uff0c\u56e0\u4e3a\u53ef\u7528\u4e8e\u518d\u5206\u6790\u7684\u8bd5\u9a8c\u901a\u5e38\u88ab\u8bbe\u8ba1\u4e3a\u4ea7\u751f\u5e73\u5747\u6548\u5e94( 116 )\u3002\u603b\u7684\u6765\u8bf4\uff0c\u6211\u4eec\u5efa\u8bae\u5927\u591a\u6570\u4f7f\u7528ML\u8bc6\u522b\u5f02\u8d28\u6027\u5904\u7406\u6548\u5e94\u7684\u65b9\u6cd5\u7528\u4e8e\u5047\u8bbe\u751f\u6210\uff0c\u5e76\u5728\u5916\u90e8\u4eba\u7fa4\u4e2d\u9a8c\u8bc1\u7279\u5b9a\u7684\u4e9a\u7ec4\u6548\u5e94\u3002<\/span><\/section>\n

<\/span><\/section>\n

\n
6.1.5 \u8bc4\u4f30\u548c\u6d88\u9664\u504f\u5dee<\/span><\/strong><\/section>\n

<\/span><\/section>\n<\/h4>\n
\u6700\u540e\uff0c\u4e00\u4e9b\u6709\u8da3\u7684ML\u5e94\u7528\u5df2\u7ecf\u88ab\u5f00\u53d1\u51fa\u6765\uff0c\u4ee5\u91cf\u5316\u548c\u89e3\u51b3\u9488\u5bf9\u56e0\u679c\u63a8\u65ad\u7684\u5206\u6790\u4e2d\u7684\u6f5c\u5728\u504f\u5dee\u3002\u5728\u7f3a\u4e4f\u53ef\u5ffd\u7565\u6027<\/span>\uff08\u65e0\u4e0d\u53ef\u6d4b\u91cf\u7684\u6df7\u6742\u56e0\u7d20\uff09<\/span>\u7684\u60c5\u51b5\u4e0b\uff0c\u5bf9\u4e0d\u53ef\u6d4b\u91cf\u6df7\u6742\u7684\u654f\u611f\u6027\u53ef\u80fd\u4f1a\u4e25\u91cd\u9650\u5236\u7814\u7a76\u7ed3\u679c\u7684\u53ef\u63a8\u5e7f\u6027\u3002\u201ctreatsens\u201d\u7a0b\u5e8f\u5305\u4f30\u8ba1\u4e86\u4e00\u4e2a\u4e0d\u53ef\u6d4b\u91cf\u7684\u6df7\u6742\u56e0\u7d20\u7684\u5927\u5c0f\uff0c\u8fd9\u5bf9\u4e8e\u6d88\u9664\u6cbb\u7597\u4e0e\u7ed3\u679c\u4e4b\u95f4\u7684\u8054\u7cfb\u662f\u5fc5\u8981\u7684\uff1b\u7136\u800c\uff0c\u8fd9\u79cd\u5206\u6790\u9700\u8981\u9886\u57df\u77e5\u8bc6(56)\u3002\u5728<\/span>\uff08\u6216\u8005\u5b9e\u73b0\u4e0d\u5f53\uff09<\/span>\u968f\u673a\u5316\u6709\u9650\u7684\u60c5\u51b5\u4e0b\uff0c\u534f\u53d8\u91cf\u7684\u4e0d\u5e73\u8861\u5206\u5e03\u53ef\u80fd\u4f1a\u4f7f\u7814\u7a76\u7ed3\u679c\u4ea7\u751f\u504f\u5dee\u3002\u5728\u8fd9\u91cc\uff0cBART\u53ef\u4ee5\u8bc4\u4f30\u7f3a\u4e4f\u5171\u540c\u652f\u6301(117)\uff0c\u800c\u534f\u53d8\u91cf\u4f18\u5148\u4e0e\u5339\u914d\u53ef\u4ee5\u8c03\u6574\u63a5\u53d7\u6cbb\u7597\u7684\u5dee\u5f02\u6982\u7387(118)\u3002\u5173\u4e8eML\u7528\u4e8e\u9006\u6982\u7387\u52a0\u6743\u548c\u503e\u5411\u5f97\u5206\u5339\u914d\u7684\u4e0d\u540c\u6587\u732e\u7684\u7efc\u8ff0\u89c1(102)\u3002\u7279\u522b\u662f\u5728\u8870\u8001\u7814\u7a76\u65b9\u9762\uff0c\u6211\u4eec\u5efa\u8bae\u7cfb\u7edf\u8bc4\u4f30\u9009\u62e9\u6027\u78e8\u8680\u548c\u7ade\u4e89\u6b7b\u4ea1\u98ce\u9669\u5e26\u6765\u7684\u504f\u501a\uff0c\u5982\u968f\u673a\u751f\u5b58\u68ee\u6797(119)\u5df2\u5e94\u7528\u4e8e\u75f4\u5446\u7814\u7a76(41)\u3002<\/span><\/section>\n

<\/span><\/section>\n

\n
6.2 \u5e94\u75282\uff1a\u56e0\u679c\u53d1\u73b0\u6216\u56e0\u679c\u7ed3\u6784\u5b66\u4e60<\/span><\/strong><\/span><\/section>\n

<\/span><\/section>\n<\/h3>\n
\u4e0eHern\u00e1n\u7b49(6)\u7684\u6846\u67b6\u4e0d\u540c\uff0c\u5176\u4ed6\u65b9\u6cd5\u8ba4\u4e3a\u56e0\u679c\u63a8\u65ad\u4e0d\u4e00\u5b9a\u9700\u8981\u53cd\u4e8b\u5b9e\u9884\u6d4b(120)\u3002\u867d\u7136\u5728\u536b\u751f\u548c\u793e\u4f1a\u79d1\u5b66\u4e2d\u5f88\u5c11\u4f7f\u7528\uff0c\u4f46\u4ece\u6570\u636e\u4e2d\u5b66\u4e60\u56e0\u679c\u7ed3\u6784\u5728\u53ef\u80fd\u4ea7\u751f<\/span>\uff08\u6cbb\u7597\u624b\u6cd5\uff09<\/span>\u7684\u60c5\u51b5\u4e0b\u7279\u522b\u6709\u8da3\u3002\u7136\u800c\uff0c\u5373\u4f7f\u5728\u4e0d\u53ef\u80fd\u8fdb\u884c\u6cbb\u7597\u64cd\u7eb5\u7684\u60c5\u51b5\u4e0b\uff0c\u4f8b\u5982\u5728\u6709\u89c2\u5bdf\u6570\u636e\u7684\u60c5\u51b5\u4e0b\uff0c\u56e0\u679c\u7ed3\u6784\u5b66\u4e60\u4e5f\u53ef\u4ee5\u9610\u660e\u56e0\u679c\u7814\u7a76\u95ee\u9898\uff1a\u90a3\u4e48\u5728<\/span>\uff08\u771f\u5b9e\u7684\u6216\u5047\u8bbe\u7684\uff09<\/span>\u5e72\u9884\u4e2d\u968f\u673a\u5206\u914d\u66b4\u9732\u7684\u610f\u4e49\u4e0a\u4e0d\u80fd\u88ab\u64cd\u7eb5\u7684\u63a8\u5b9a\u539f\u56e0\u5462? \u6211\u4eec\u8ba4\u4e3a\uff0c\u5047\u60f3\u5e72\u9884\u7684\u7f3a\u5931\u4e0d\u5e94\u8be5\u9650\u5236\u6211\u4eec\u4f30\u8ba1\u56e0\u679c\u6548\u5e94\uff0c\u5982\u6027\u522b\uff0c\u56e0\u4e3a\u8bc4\u4f30\u793e\u4f1a\u56e0\u7d20\u7684\u6548\u5e94\u5927\u5c0f\u662f\u63d0\u9ad8\u6211\u4eec\u5bf9\u6240\u7814\u7a76\u73b0\u8c61\u7684\u7406\u89e3\u548c\u5236\u5b9a\u6709\u9488\u5bf9\u6027\u7684\u5e72\u9884\u63aa\u65bd\u7684\u5148\u51b3\u6761\u4ef6(121)\u3002\u79cd\u65cf\u548c\u6027\u522b\/\u6027\u522b\u4e0e\u793e\u4f1a\u7ecf\u6d4e\u548c\u884c\u4e3a\u6307\u6807\u7684\u4e0d\u540c\u5173\u7cfb\u80fd\u591f\u6df1\u5165\u4e86\u89e3\u793e\u4f1a\u72b6\u51b5(122)\u3002\u79cd\u65cf\u6216\u6027\u522b\/\u6027\u522b\u7b49\u4e0d\u53ef\u64cd\u4f5c\u7684\u53d8\u91cf\u53ef\u4ee5\u901a\u8fc7\u5141\u8bb8\u64cd\u7eb5\u6216\u67d0\u79cd\u5f62\u5f0f\u5e72\u9884\u7684\u65b9\u5f0f\u91cd\u65b0\u6982\u5ff5\u5316\uff0c\u4f8b\u5982\u5728\u7ef4\u6d85\u7279\u7814\u7a76\u4e2d\u4f7f\u7528\u611f\u77e5\u79cd\u65cf\u3002\u7136\u800c\uff0c\u8d85\u8d8a\u6f5c\u5728\u7ed3\u679c\u6846\u67b6\uff0c\u6211\u4eec\u8ba4\u4e3a\uff0c\u5728\u6ca1\u6709\u5047\u8bbe\u5e72\u9884\u7684\u60c5\u51b5\u4e0b\uff0c\u6211\u4eec\u53ef\u4ee5\u5229\u7528\u7ed3\u6784\u56e0\u679c\u5b66\u4e60\uff0c\u5373\u4f7f\u7528\u80fd\u591f\u4ece\u6570\u636e\u4e2d\u5b66\u4e60<\/span>\uff08\u548c\u5448\u73b0\uff09<\/span>\u53d8\u91cf\u7684\u56e0\u679c\u7ed3\u6784\u7684\u7b97\u6cd5\u3002\u6700\u8fd1\u63d0\u51fa\u4e86\u961f\u5217\u6570\u636e\u4e2d\u56e0\u679c\u53d1\u73b0\u65b9\u6cd5\u7684\u6307\u5357(123)\u3002\u5728\u4e0b\u6587\u4e2d\uff0c\u6211\u4eec\u5c06\u9996\u5148\u5bf9\u7ed3\u6784\u5b66\u4e60\u8fdb\u884c\u6982\u8ff0\uff0c\u7136\u540e\u8f6c\u5411\u4ece\u56fe\u5f62\u7ed3\u6784\u63a8\u5bfc\u56e0\u679c\u63a8\u7406\u8fd9\u4e00\u66f4\u4e3a\u590d\u6742\u7684\u4efb\u52a1\uff0c\u5373\u6240\u8c13\u7684\u56e0\u679c\u7ed3\u6784\u5b66\u4e60\u3002<\/span><\/section>\n

<\/span><\/section>\n

\n
6.2.1 \u7ed3\u6784\u5b66\u4e60<\/span><\/strong><\/section>\n

<\/span><\/section>\n<\/h4>\n
\u6211\u4eec\u9996\u5148\u8003\u8651\u65e0\u5411\u56fe\u7684\u7ed3\u6784\u5b66\u4e60\uff0c\u5373\u5b66\u4e60\u8de8\u8d8a\u590d\u6742\u9ad8\u7ef4\u6570\u636e\u7684\u6761\u4ef6\u72ec\u7acb\u7ed3\u6784\u3002\u56fe\u5f62LASSO (124)\u5728\u6b64\u8bbe\u7f6e\u4e2d\u88ab\u5e7f\u6cdb\u91c7\u7528\uff0c\u5e76\u4e14\u57fa\u4e8e\u4e00\u4e2a\u6f5c\u5728\u7684\u9ad8\u65af\u5047\u8bbe<\/span>\uff08\u5373\u5bf9\u4e8e\u8fde\u7eed\u53d8\u91cf\uff09<\/span>\u3002\u9ad8\u65af\u56fe\u5f62LASSO\u7684\u5404\u79cd\u6269\u5c55\u5df2\u7ecf\u88ab\u5f00\u53d1\u51fa\u6765\uff0c\u5305\u62ec\u540c\u65f6\u5b66\u4e60\u548c\u4f30\u8ba1\u53d8\u91cf\u8de8\u7ec4<\/span>\uff08\u5982\u5bf9\u5e94\u4e0d\u540c\u7684\u4e9a\u7fa4\u6216\u4e0d\u540c\u6761\u4ef6\u4e0b\u6536\u96c6\u7684\u6570\u636e\uff09<\/span>\u548c\u8de8\u7a7a\u95f4\u548c\/\u6216\u65f6\u95f4<\/span>\uff08\u5982\u7eb5\u5411\u6570\u636e\uff09<\/span>\u7684\u7f51\u7edc\u7ed3\u6784\u3001\u6df7\u5408\u53d8\u91cf<\/span>\uff08\u5373\u5bf9\u8fde\u7eed\u53d8\u91cf\u548c\u79bb\u6563\u53d8\u91cf\u7684\u6d4b\u91cf\uff09<\/span>\u548c\u7f3a\u5931\u6570\u636e\u7684\u6269\u5c55\u3002<\/span><\/section>\n

<\/span><\/section>\n
\u8fd9\u5728\u6570\u636e\u7a00\u758f\u7684\u60c5\u51b5\u4e0b\u5c24\u5176\u6709\u8da3\uff0c\u5373\u5305\u542b\u8bb8\u591a\u7a7a\u5355\u5143\u683c\uff0c\u540c\u65f6\u5904\u7406\u6240\u6709\u53ef\u7528\u6570\u636e\u5c06\u4ed8\u51fa\u6781\u9ad8\u7684\u8ba1\u7b97\u4ee3\u4ef7\u3002\u6700\u8fd1\u7684\u4e00\u9879\u7814\u7a76\u4f7f\u7528\u6df7\u5408\u53d8\u91cf\uff0c\u5373\u4e8c\u5143\u3001\u8fde\u7eed\u548c\u8ba1\u6570\u53d8\u91cf\u7684\u6f5c\u5728\u9ad8\u65af\u56fe\u5f62\u6a21\u578b\u6765\u63a8\u65ad\u53e3\u5934\u5c38\u68c0\u4e2d\u7684\u75c7\u72b6\u5173\u8054(125)\uff0c\u8fd9\u53ef\u80fd\u6709\u52a9\u4e8e\u5f97\u51fa\u66f4\u7a33\u5065\u7684\u53ef\u80fd\u6b7b\u4ea1\u539f\u56e0\u5206\u7c7b\u3002<\/span><\/section>\n

<\/span><\/section>\n

\n
6.2.2 \u56e0\u679c\u7ed3\u6784\u5b66\u4e60<\/span><\/strong><\/section>\n

<\/span><\/section>\n<\/h4>\n
\u56e0\u679c\u7ed3\u6784\u5b66\u4e60\u901a\u8fc7\u63a8\u65ad\u56fe\u4e2d\u8fb9\u7684\u65b9\u5411\u6765\u6269\u5c55\u7ed3\u6784\u5b66\u4e60\u3002\u57fa\u4e8e\u7ea6\u675f\u7684\u65b9\u6cd5\u9996\u5148\u5b66\u4e60\u4e00\u4e2a\u65e0\u5411\u56fe\uff0c\u8868\u793aDAG\u7684\u9aa8\u67b6\uff0c\u7136\u540e\u786e\u5b9a\u65b9\u5411\u3002\u6216\u8005\uff0c\u57fa\u4e8e\u5f97\u5206\u7684\u65b9\u6cd5\u76f4\u63a5\u5728DAGs\u7684\u7a7a\u95f4\u4e0a\u641c\u7d22\uff0c\u5e76\u6839\u636e\u6307\u5b9a\u7684\u76ee\u6807\u51fd\u6570\u5bf9\u6bcf\u4e2a\u56fe\u8fdb\u884c\u8bc4\u5206\uff1b\u968f\u7740\u8282\u70b9\/\u53d8\u91cf\u6570\u91cf\u5448\u8d85\u6307\u6570\u589e\u957f\u7684DAG\u6570\u91cf\u5e9e\u5927\uff0c\u9700\u8981\u7cbe\u5fc3\u8bbe\u8ba1\u641c\u7d22\u7b97\u6cd5\u548c\u8bc4\u5206\u3002\u6df7\u5408\u7b97\u6cd5\u7ed3\u5408\u4e86\u57fa\u4e8e\u7ea6\u675f\u548c\u57fa\u4e8e\u5206\u6570\u7684\u65b9\u6cd5\u7684\u601d\u60f3\u3002Pearl (120)\u5bf9\u56e0\u679c\u5173\u7cfb\u8fdb\u884c\u4e86\u5168\u9762\u7684\u6280\u672f\u6982\u8ff0\uff0c\u6700\u8fd1\u7684\u56e0\u679c\u7ed3\u6784\u5b66\u4e60\u7efc\u8ff0\u5728(126\u3001127)\u4e2d\u63d0\u4f9b\u3002\u566a\u58f0\u6570\u636e\u5047\u8bbe\u4e0b\u56e0\u679c\u7ed3\u6784\u5b66\u4e60\u7b97\u6cd5\u7684\u5b9e\u8bc1\u8bc4\u4f30\u548c\u6bd4\u8f83\u89c1(128)\u3002\u5728\u56e0\u679c\u7ed3\u6784\u5b66\u4e60\u7684\u5e94\u7528\u4e2d\uff0c\u6027\u80fd\u4f9d\u8d56\u4e8e\u6f5c\u5728\u5047\u8bbe<\/span>\uff08\u4e0d\u540c\u7684\u6a21\u578b\u4f9d\u8d56\u4e8e\u4e0d\u540c\u7684\u5047\u8bbe\uff09<\/span>\u7684\u9002\u5f53\u6027\uff0c\u6837\u672c\u91cf\u5bf9\u6027\u80fd<\/span>\uff08\u4ece\u51e0\u767e\u5230\u4e24\u4e07\u4e0d\u7b49\uff09<\/span>\u7684\u5f71\u54cd\u5f88\u5c0f\uff0c\u66f4\u7a00\u758f\u7684\u56fe\u66f4\u5bb9\u6613\u4f30\u8ba1(127)\u3002R\u4e2d\u7684BNLEARN\u5305\u5b9e\u73b0\u4e86\u591a\u4e2a\u57fa\u4e8e\u7ea6\u675f\u548c\u57fa\u4e8e\u5f97\u5206\u7684\u56e0\u679c\u7ed3\u6784\u5b66\u4e60\u7b97\u6cd5(129)\u3002\u57fa\u4e8e\u7ea6\u675f\u7684\u65b9\u6cd5\uff0c\u7279\u522b\u662fBNLEARN\u4e2d\u7684\u589e\u91cf\u5173\u8054\u9a6c\u5c14\u79d1\u592b\u6bef\u7b97\u6cd5\uff0c\u5df2\u88ab\u7528\u4e8e\u8bc6\u522b\u6492\u54c8\u62c9\u4ee5\u5357\u975e\u6d3229\u4e2a\u56fd\u5bb6<\/span>\uff08130\u4e2a\uff09<\/span>\u4eba\u53e3\u548c\u5065\u5eb7\u8c03\u67e5\u4e2d\u7684HIV\u72b6\u6001\u3001\u68c0\u6d4b\u548c\u77e5\u8bc6\u7684\u98ce\u9669\u56e0\u7d20<\/span>\uff08\u8fd9\u53ef\u80fd\u662f\u5bfc\u81f4\uff09<\/span>\u3002\u867d\u7136\u6570\u636e\u662f\u6a2a\u622a\u9762\u7684\uff0c\u6ca1\u6709\u4e0d\u53ef\u6d4b\u91cf\u7684\u6df7\u6742\uff0c\u4f46\u7f3a\u4e4fHIV\u72b6\u51b5\u7684\u793e\u4f1a\u884c\u4e3a\u51b3\u5b9a\u56e0\u7d20\uff0c\u4f46\u5b58\u5728HIV\u77e5\u8bc6\u548c\u68c0\u6d4b\u7684\u793e\u4f1a\u884c\u4e3a\u51b3\u5b9a\u56e0\u7d20\uff0c\u8fd9\u5c31\u4f7f\u6211\u4eec\u4e86\u89e3\u4e86\u5728\u786e\u5b9a\u4f20\u67d3\u75c5\u4f20\u64ad\u65f6\u8c03\u67e5\u6570\u636e\u6536\u96c6\u7684\u5c40\u9650\u6027\u3002\u53e6\u4e00\u9879\u7814\u7a76\u4f7f\u7528Markov\u548c\u57fa\u4e8e\u5206\u6570\u7684\u5feb\u901f\u8d2a\u5a6a\u7b49\u4ef7\u641c\u7d22\u6765\u8bc6\u522b\u79cd\u65cf\u548c\u6027\u522b\/\u6027\u522b\u7b49\u6784\u6210\u56e0\u7d20(122)\u3002\u867d\u7136\u6027\u522b\u4e3b\u8981\u4e0e\u4eba\u683c\u7279\u5f81\u548c\u884c\u4e3a\u6709\u5173\uff0c\u4f46\u79cd\u65cf\u4e0e\u793e\u4f1a\u7ecf\u6d4e\u5265\u593a\u7684\u7cfb\u7edf\u6027\u56e0\u7d20\u548c\u884c\u4e3a\u6709\u5173\uff0c\u8fd9\u4e0e\u79cd\u65cf\u7c7b\u522b\u662f\u5728\u4e00\u4e2a\u7cfb\u7edf<\/span>\uff08\u7ed3\u6784\u6027\u79cd\u65cf\u4e3b\u4e49\uff09<\/span>(122)\u4e2d\u6784\u5efa\u7684\u89c2\u70b9\u4e00\u81f4\u3002<\/span><\/section>\n

<\/span><\/section>\n
\u6211\u4eec\u8ba4\u4e3a\u8fd9\u4e9b\u65b9\u6cd5\u7684\u5e94\u7528\u9700\u8981\u7ecf\u8fc7\u8fd9\u4e9b\u65b9\u6cd5\u57f9\u8bad\u7684\u5408\u4f5c\u8005\u5bf9\u793e\u4f1a\u548c\u5065\u5eb7\u7814\u7a76\u4eba\u5458\u8fdb\u884c\u57f9\u8bad\uff0c\u56e0\u4e3a\u4f7f\u7528\u7ed3\u6784\u56e0\u679c\u5b66\u4e60\u65b9\u6cd5\u4e5f\u610f\u5473\u7740\u5fc5\u987b\u52aa\u529b\u7406\u89e3\u548c\u4ea4\u6d41\u8c03\u67e5\u7ed3\u679c\u3002\u4f7f\u7528\u8fd9\u79cd\u65b9\u6cd5\u5c06\u4ea7\u751f\u66f4\u590d\u6742\u7684\u7ed3\u679c\u96c6\uff0c\u4f8b\u5982\u8d8a\u6765\u8d8a\u591a\u7684\u53ef\u80fd\u7ade\u4e89\u7684\u56fe\u3002\u968f\u7740\u5bf9\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u4e2d\u51e0\u4e4e\u6240\u6709\u5b66\u79d1\u7684\u6f5c\u5728\u7ed3\u679c\u6846\u67b6\u7684\u5f3a\u70c8\u5f3a\u8c03\uff0c\u5176\u4ed6\u56e0\u679c\u63a8\u65ad\u65b9\u6cd5<\/span>\uff08\u53ca\u5176\u6f5c\u5728\u5229\u76ca\uff09<\/span>\u5728\u8fd9\u4e2a\u65f6\u5019\u4ecd\u7136\u662f\u672a\u77e5\u7684\u3002\u5728\u8ba8\u8bba\u4e86\u7528\u4e8e\u63cf\u8ff0\u3001\u9884\u6d4b\u548c\u56e0\u679c\u63a8\u7406\u7684ML\u4efb\u52a1\u4e4b\u540e\uff0c\u6211\u4eec\u5c06\u5728\u4e0b\u4e00\u8282\u8be6\u7ec6\u4ecb\u7ecd\u6700\u5e38\u89c1\u7684ML\u6027\u80fd\u6307\u6807\u3002<\/span><\/section>\n

\n

<\/span><\/section>\n


<\/p>\n<\/h2>\n

\n
\n
\n
\n
\n

<\/section>\n

<\/section>\n
\n
<\/svg><\/section>\n<\/section>\n<\/section>\n
\n

7. \u673a\u5668\u5b66\u4e60\u8868\u73b0\u6027\u80fd\u8bc4\u4f30<\/strong><\/strong><\/p>\n<\/section>\n

\n

<\/section>\n

<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/h3>\n

<\/section>\n
\u5728\u5c06ML\u65b9\u6cd5\u6620\u5c04\u5230\u63cf\u8ff0\u3001\u9884\u6d4b\u548c\u56e0\u679c\u63a8\u7406\u7684\u7814\u7a76\u76ee\u6807\u4e4b\u540e\uff0c\u53ef\u4ee5\u6bd4\u8f83\u4e0d\u540cML\u6a21\u578b\u7684\u6027\u80fd\u6216\u901a\u8fc7\u4e0e\u66f4\u4f20\u7edf\u7684\u7edf\u8ba1\u65b9\u6cd5\u76f8\u6bd4\u8bc4\u4f30ML\u3002\u4e3a\u4e86\u8fd9\u4e2a\u76ee\u7684\uff0c\u88684\u7ed9\u51fa\u4e86\u4e00\u4e2a\u975e\u8be6\u5c3d\u7684\u5e38\u7528\u7684\u8bc4\u4f30ML\u6a21\u578b\u6027\u80fd\u6307\u6807\u5217\u8868\u3002\u901a\u8fc7\u8bc4\u4f30\u6a21\u578b\u7684\u8d28\u91cf\u548c\u6027\u80fd\uff0cML\u6a21\u578b\u76f8\u6bd4\u4e8e\u4f20\u7edf\u7684\u5efa\u6a21\u80fd\u591f\u5728\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u9886\u57df\u4e2d\u63d0\u4f9b\u95f4\u63a5\u7684\u89e3\u51b3\u65b9\u6848\u3002\u4f8b\u5982\uff0c\u5728\u7ecf\u6d4e\u5b66\u4e2d\u9700\u8981\u6ee1\u8db3\u5e38\u6001\u6027\u3001\u4e00\u81f4\u6027\u548c\u6548\u7387\u7684\u6761\u4ef6\uff0c\u5e76\u4e14\u53ef\u4ee5\u8fdb\u884c\u76f8\u5f53\u7684\u76f4\u63a5\u7684\u6d4b\u8bd5\uff08113\uff09\u3002<\/span><\/section>\n

<\/span><\/section>\n
\u88684 \u7efc\u8ff0\u4e0e\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u4e2d\u6700\u76f8\u5173\u7684\u673a\u5668\u5b66\u4e60\u65b9\u6cd5\u975e\u6280\u672f\u63cf\u8ff0<\/span>
<\/section>\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n
\n
\u6307\u6807<\/span><\/strong><\/section>\n<\/td>\n
\n
\u89e3\u91ca\u548c\u4f8b\u5b50<\/span><\/strong><\/section>\n<\/td>\n<\/tr>\n
\n
\u65e0\u76d1\u7763\u5b66\u4e60<\/span><\/strong><\/section>\n<\/td>\n<\/tr>\n
\n
\u805a\u7c7b\u76f8\u5173<\/span><\/strong><\/section>\n<\/td>\n<\/tr>\n
\n
ARI\u6307\u6570<\/span><\/section>\n<\/td>\n
\n
\u5ea6\u91cf\u4e24\u4e2a\u6570\u636e\u805a\u7c7b\u7684\u76f8\u4f3c\u6027\uff1b\u4e0e\u7cbe\u5ea6\u6709\u5173\uff0c\u4f46\u5bf9\u4e8e\u65e0\u6807\u7b7e\u6570\u636e\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
\u4e92\u4fe1\u606f\u71b5<\/span><\/section>\n<\/td>\n
\n
\u5ea6\u91cf\u76f8\u4e92\u4f9d\u8d56\u7a0b\u5ea6\uff0c\u8bc4\u4ef7\u4e24\u7ec4\u53d8\u91cf\u7684\u8054\u5408\u5206\u5e03\u5bf9\u5404\u81ea\u8fb9\u7f18\u5206\u5e03\u4e58\u79ef\u7684\u5dee\u5f02\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
Calinski-Harabasz<\/span><\/section>\n
\u6307\u6570<\/span><\/section>\n<\/td>\n
\n
\u786e\u5b9a\u201c\u6b63\u786e\u201d\u805a\u7c7b\u6570\u76ee\u7684\u6807\u51c6\uff0c\u51e0\u4e2a\u76f8\u5173\u7684\u6807\u51c6\uff0c\u4f8b\u5982\uff0c\u5728kml\u5305\u4e2d\u5b9e\u73b0\u5bf9\u7eb5\u5411\u6570\u636e\u7684\u805a\u7c7b<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
Dunn\u6307\u6570<\/span><\/section>\n<\/td>\n
\n
\u9a8c\u8bc1\u805a\u7c7b\u89e3<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
\u964d\u7ef4\u76f8\u5173<\/span><\/strong><\/section>\n<\/td>\n<\/tr>\n
\n
\u91cd\u6784\u8bef\u5dee<\/span><\/section>\n<\/td>\n
\n
\u4ece\u63a8\u65ad\u7684\u4f4e\u7ef4\u6f5c\u53d8\u91cf\u4e2d\u6d4b\u91cf\u89c2\u6d4b\u6570\u636e\u548c”\u91cd\u5efa\u6570\u636e”\u4e4b\u95f4\u7684\u8ddd\u79bb(\u4f8b\u5982,\u8fde\u7eed\u6570\u636e\u7684\u6b27\u51e0\u91cc\u5f97)\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
\u6709\u76d1\u7763\u5b66\u4e60<\/span><\/strong><\/section>\n<\/td>\n<\/tr>\n
\n
\u91cd\u8981\u6027\u6307\u6807<\/span><\/section>\n<\/td>\n
\n
\u53d8\u91cf\u91cd\u8981\u6027\u91cf\u5316\u4e86\u4e00\u4e2a\u53d8\u91cf\u5bf9\u5206\u7c7b\u6216\u56de\u5f52\u8868\u73b0\u7684\u4e2a\u4f53\u8d21\u732e\u3002\u6709\u51e0\u79cd\u5b9e\u73b0\u65b9\u5f0f\u3002\u5bf9\u4e8e\u57fa\u4e8e\u6811\u7684\u6a21\u578b\uff0c\u5982\u968f\u673a\u68ee\u6797\uff0c\u53d8\u91cf\u91cd\u8981\u6027\u901a\u5e38\u88ab\u5efa\u6a21\u4e3a\u4f7f\u7528\u53d8\u91cf\u5728\u6240\u6709\u6811\u4e0a\u5e73\u5747\u5f97\u5230\u7684\u6539\u8fdb\u4e4b\u548c( 4 )\u3002\u5728\u5206\u7c7b\u548c\u56de\u5f52\u4e2d\uff0c\u901a\u5e385\u523010\u4e2a\u6700\u91cd\u8981\u7684\u53d8\u91cf\u53ef\u4ee5\u88ab\u6709\u610f\u4e49\u5730\u89e3\u91ca\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
\u7531\u4e8e\u91cd\u8981\u6027\u4e0d\u5bb9\u6613\u4e0e\u4f20\u7edf\u7684\u7edf\u8ba1\u5b66\u6307\u6807\u76f8\u6bd4\u8f83\uff0c\u7814\u7a76\u4eba\u5458\u53ef\u4ee5\u5728\u591a\u4e2a\u6a21\u578b\u4e2d\u6bd4\u8f83\u53d8\u91cf\u91cd\u8981\u6027\uff0c\u5e76\u8003\u8651\u5230\u7edf\u8ba1\u5b66\u548c\u4e34\u5e8a(\u5e94\u7528)\u91cd\u8981\u6027\uff0c\u589e\u52a0\u4e00\u4e2a\u968f\u673a\u53d8\u91cf\u4f5c\u4e3a\u57fa\u51c6\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
\u56de\u5f52\u76f8\u5173<\/span><\/strong><\/section>\n<\/td>\n<\/tr>\n
\n
\u51c6\u786e\u7387<\/span><\/section>\n<\/td>\n
\n
\u5728\u6240\u6709\u9884\u6d4b( =\u771f\u9633\u6027+\u771f\u9634\u6027\/\u771f\u9633\u6027+\u771f\u9634\u6027+\u5047\u9633\u6027+\u5047\u9634\u6027)\u4e0a\u6b63\u786e\u5206\u7c7b\u5b9e\u4f8b\u7684\u6bd4\u7387\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
\u5728(\u63a5\u8fd1)\u5e73\u8861\u6570\u636e\u4e2d\u7684\u826f\u597d\u6d4b\u91cf\uff0c\u5373\u76f8\u4f3c\u7387\u7684\u7ed3\u679c\u7c7b\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
\u4e0d\u8981\u5728\u4e0d\u5e73\u8861\u6570\u636e\u4e2d\u4f7f\u7528\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
\u5747\u8861\u51c6\u786e\u7387<\/span><\/section>\n<\/td>\n
\n
\u7075\u654f\u5ea6\u548c\u7279\u5f02\u5ea6\u7684\u7b97\u672f\u5e73\u5747\u503c(\u89c1\u4e0b\u6587)\uff1b\u5e73\u5747\u6bcf\u4e2a\u7c7b\u522b\u7684\u6b63\u786e\u5206\u7c7b\u6848\u4f8b\u7684\u6bd4\u4f8b\uff0c\u5728\u4e0d\u5e73\u8861\u7c7b\u4e2d\u76f8\u5173\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
\u5747\u65b9\u6839\u8bef\u5dee<\/span><\/section>\n<\/td>\n
\n
\u76ee\u6807\u503c\u4e0e\u6a21\u578b\u9884\u6d4b\u503c\u4e4b\u95f4\u7684\u5e73\u5747\u5e73\u65b9\u5dee\u3002\u60e9\u7f5a\u5927\u7684\u9519\u8bef\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
\u5e73\u65b9\u6839\u8bef\u5dee<\/span><\/section>\n<\/td>\n
\n
\u56de\u5f52\u4efb\u52a1\u7684\u9996\u9009\u5ea6\u91cf\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
\u539f\u59cb\u503c\u4e0e\u9884\u6d4b\u503c\u4e4b\u5dee\u7684\u5e73\u65b9\u7684\u5e73\u5747\u503c\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
\u6392\u5217\u7b49\u7ea7<\/span><\/section>\n<\/td>\n
\n
\u8bc4\u4f30\u8de8\u6a21\u578b\u6216\u8de8\u6837\u672c\u9884\u6d4b\u56e0\u5b50\u7684\u91cd\u8981\u6027\uff0c\u5982\u9884\u6d4b\u75f4\u5446(41)\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
\u98ce\u9669\u6bd4\u7387<\/span><\/section>\n<\/td>\n
\n
\u5728\u56de\u5f52\u4e2d\u8bc4\u4f30\u5355\u4e2a\u9884\u6d4b\u56e0\u5b50\uff0c\u5982\u9884\u6d4b\u75f4\u5446(41)\uff0c\u53ef\u4ee5\u7528\u7f6e\u4fe1\u533a\u95f4\u62a5\u544a\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
\u7edd\u5bf9\u5e73\u5747\u8bef\u5dee<\/span><\/section>\n<\/td>\n
\n
\u539f\u59cb\u503c\u4e0e\u9884\u6d4b\u503c\u4e4b\u95f4\u7edd\u5bf9\u5dee\u503c\u7684\u5e73\u5747\u503c\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
\u5206\u7c7b\u76f8\u5173<\/span><\/strong><\/section>\n<\/td>\n<\/tr>\n
\n
AUC<\/span><\/section>\n<\/td>\n
\n
\u5bf9\u4e8e\u4e8c\u5206\u7c7b\u7ed3\u679c\u7684\u5206\u7c7b\uff0c\u8be5\u6307\u6807\u89c4\u5b9a\u4e86ROC\u66f2\u7ebf\u4e0b\u7684\u9762\u79ef\uff0c\u89c1\u4e0b\u6587\uff0c\u8303\u56f4\u57280\u548c1\u4e4b\u95f4\u3002\u6570\u503c\u8d8a\u5927\uff0c\u6a21\u578b\u8d8a\u597d\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
\u5bf9\u6570\u635f\u5931<\/span><\/section>\n<\/td>\n
\n
\u5ea6\u91cf\u4e00\u4e2a\u5206\u7c7b\u6a21\u578b\u7684\u6027\u80fd\uff0c\u8be5\u6a21\u578b\u63d0\u4f9b\u9884\u6d4b\u7684\u7c7b\u6982\u7387\u3002\u5f53\u9884\u6d4b\u6982\u7387\u4e0e\u5b9e\u9645\u7c7b\u522b\u503c\u7684\u504f\u5dee\u589e\u5927\u65f6\uff0c\u65e5\u5fd7\u635f\u5931\u4f1a\u53d8\u5927\u3002\u60e9\u7f5a\u9519\u8bef\u5206\u7c7b\u3002\u5728\u591a\u7c7b\u5206\u7c7b(\u89c1\u591a\u9879\u5f0f\u5bf9\u6570\u635f\u5931)\u4e2d\u6548\u679c\u826f\u597d\u3002\u6700\u5c0f\u5316\u65e5\u5fd7\u635f\u5931\u4f1a\u5f97\u5230\u66f4\u9ad8\u7684\u51c6\u786e\u7387\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
\u5e73\u5747\u7edd\u5bf9\u767e\u5206\u6bd4<\/span><\/section>\n<\/td>\n
\n
\u9884\u6d4b\u7cbe\u5ea6\u7684\u5ea6\u91cf\u3002\u5b83\u901a\u5e38\u5c06\u51c6\u786e\u7387\u8868\u793a\u4e3a\u4e00\u4e2a\u6bd4\u7387\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
\u7cbe\u5ea6<\/span><\/section>\n<\/td>\n
\n
\u5728\u4e8c\u5206\u7c7b\u4e2d\uff0c\u5f53\u8bef\u62a5\u4ee3\u4ef7\u8f83\u9ad8\u65f6\u76f8\u5173\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
\u6b63\u786e\u5206\u7c7b\u75c5\u4f8b\u5360\u5168\u90e8\u5206\u7c7b\u75c5\u4f8b\u7684\u6bd4\u4f8b(\u4f8b\u5982,\u88ab\u8bd5)\uff0c\u767e\u5206\u6bd4\uff0c=\u771f\u9633\u6027\/ (\u771f\u9633\u6027+\u5047\u9633\u6027)\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
\u6b63\u786e\u7387<\/span><\/section>\n<\/td>\n
\n
\u9002\u7528\u4e8e\u591a\u7c7b\u5206\u7c7b(\u5177\u6709\u4e24\u4e2a\u4ee5\u4e0a\u7c7b\u522b)\uff0c\u6613\u4e8e\u89e3\u91ca\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
\u53ec\u56de\u7387<\/span><\/section>\n<\/td>\n
\n
\u5728\u5206\u7c7b\u65f6\uff0c\u5f53\u6f0f\u62a5\u4ee3\u4ef7\u8f83\u9ad8\u65f6\uff0c\u76f8\u5173\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
\u6240\u6709\u5b9e\u9645\u9633\u6027\u75c5\u4f8b\u6b63\u786e\u5206\u7c7b\u7684\u6bd4\u7387[ (\u771f\u9633\u6027+\u5047\u9634\u6027)\u771f\u9633\u6027]\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
\u6b63\u786e\u5212\u5206\u4e0d\u540c\u9636\u5c42\u7684\u75c5\u4f8b\uff0c\u5982\u98ce\u9669\u6700\u9ad8\u7684\u9636\u5c42\u3001\u98ce\u9669\u4e2d\u7b49\u7684\u9636\u5c42\u3001\u98ce\u9669\u6700\u4f4e\u7684\u9636\u5c42\uff0c\u5982\u9884\u6d4b\u8001\u5175\u81ea\u6740(71)\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
ROC\u66f2\u7ebf<\/span><\/section>\n<\/td>\n
\n
\u5728\u4e8c\u5206\u7c7b\u4e2d\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
\u5141\u8bb8\u53ef\u89c6\u5316\u771f\u9633\u6027\u7387\u548c\u5047\u9633\u6027\u7387\u4e4b\u95f4\u7684\u6743\u8861\u3002ROC\u66f2\u7ebf\u663e\u793a\u4e86\u4e00\u4e2a\u5206\u7c7b\u6a21\u578b\u5728\u6240\u6709\u5206\u7c7b\u9608\u503c\u4e0b\u7684\u6027\u80fd\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
ROC\u9762\u79ef<\/span><\/section>\n<\/td>\n
\n
\u5728\u4e8c\u5206\u7c7b\u4e2d\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
\u5c55\u793a\u4e86\u5728\u6240\u6709\u5206\u7c7b\u9608\u503c\u4e0b\uff0c\u771f\u5b9e\u9633\u6027\u7387(\u53ec\u56de\u7387\/\u654f\u611f\u5ea6)\u4e0e\u9633\u6027\u9884\u6d4b\u503c(\u7cbe\u786e\u5ea6)\u4e4b\u95f4\u7684\u6743\u8861\u3002\u5b83\u4e0e\u6b63\u4f8b\u7684\u4f4e\u6d41\u884c\u7387\u7684\u4e0d\u5e73\u8861\u6570\u636e\u96c6\u76f8\u5173\uff0c\u5176\u4e2d\u771f\u6b63\u7684\u6b63\u4f8b\u9700\u8981\u88ab\u5f3a\u8c03\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
\u6570\u503c\u8d8a\u5927\uff0c\u6a21\u578b\u6548\u679c\u8d8a\u597d\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
\u7279\u5f02\u5ea6<\/span><\/section>\n<\/td>\n
\n
\u5206\u7c7b\u95ee\u9898\u4e2d\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
\u6b63\u786e\u5206\u7c7b\u5426\u5b9a\u7684\u6bd4\u4f8b\uff0c\u771f\u5b9e\u5426\u5b9a\u7684\u6bd4\u7387\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
\u7b49\u4e8e1\u4e2a\u5047\u9633\u6027\u7684\u6bd4\u7387\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
F1\u6307\u6570<\/span><\/section>\n<\/td>\n
\n
\u7ed3\u5408\u7cbe\u786e\u7387\u548c\u53ec\u56de\u7387\uff0c\u5373F1\u503c\u8d8a\u9ad8\uff0c\u8bef\u62a5\u7387\u548c\u6f0f\u62a5\u7387\u8d8a\u4f4e\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
\u56e0\u679c\u76f8\u5173<\/span><\/strong><\/section>\n<\/td>\n<\/tr>\n
\n
\u53cd\u4e8b\u5b9e\u9884\u6d4b<\/span><\/strong><\/section>\n<\/td>\n<\/tr>\n
\n
\u5e73\u5747\u5904\u7406\u6548\u5e94<\/span><\/section>\n<\/td>\n
\n
\u5982\u679c\u6240\u6709\u89c2\u5bdf\u90fd\u63a5\u53d7\u4e86\u6cbb\u7597\uff0c\u4e0e\u6ca1\u6709\u63a5\u53d7\u6cbb\u7597\u7684\u60c5\u51b5\u76f8\u6bd4\uff0c\u7ed3\u679c\u7684\u5dee\u5f02\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
\u7edd\u5bf9\u504f\u5dee\u4f30\u8ba1<\/span><\/section>\n<\/td>\n
\n
\u5bf9\u4e0d\u53ef\u6d4b\u91cf\u6df7\u6742\u7684\u654f\u611f\u6027\uff1btreatsens\u4e2d\uff0c\u672a\u6d4b\u91cf\u7684\u6df7\u6742\u56e0\u7d20\u7684\u4f30\u8ba1\u503c\u4f7f\u5f97\u5047\u5b9a\u539f\u56e0\u7684\u5f71\u54cd\u4e3a\u96f6[ ( 56 )\u4e2d\u5173\u4e8eU\u7684\u7cfb\u6570]\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
\u6bd4\u4f8b\u4e2d\u7684\u70b9\u4f30\u8ba1\u5dee\u5f02<\/span><\/section>\n<\/td>\n
\n
\u6bd4\u8f83\u4e24\u79cd\u5904\u7406\u7684\u6548\u679c\u4f30\u8ba1\uff0c\u4f8b\u5982BART\u548c\u7528\u4e8e\u56e0\u679c\u63a8\u65ad\u7684\u5176\u4ed6\u7b97\u6cd5\uff1b\u53c2\u89c1\u6587\u732e( 118 )\u4f8b\u5b50\uff1b\u53ef\u4ee5\u662f\u5e73\u5747\u6216\u5f02\u8d28\u6027\u5904\u7406\u6548\u5e94\u3002\u4ee595 %\u7f6e\u4fe1\u533a\u95f4\u62a5\u544a\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
\u8c03\u6574\u540e\u98ce\u9669\u5dee<\/span><\/section>\n<\/td>\n
\n
\u5019\u9009\u539f\u56e0\u7684\u6548\u679c\u8bc4\u4ef7\uff1bTMLE\u4e2d\u7684\u5173\u8054\u5e73\u5747\u5904\u7406\u6548\u5e94\uff1b\u53c2\u89c1\u6587\u732e( 108 )\u4f8b\u5b50\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
\u4ee595 %\u7f6e\u4fe1\u533a\u95f4\u62a5\u544a\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
SE<\/span><\/section>\n<\/td>\n
\n
\u7ed9\u51fa\u4e86\u5904\u7406\u6548\u5e94\u4f30\u8ba1\u7684\u7f6e\u4fe1\u5ea6\u3002\u901a\u8fc7\u8ba1\u7b97\u8fdb\u884c\u7edf\u8ba1\u63a8\u65ad\uff0c\u4f8b\u5982TMLE\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
\u56e0\u679c\u53d1\u73b0<\/span><\/strong><\/section>\n<\/td>\n<\/tr>\n
\n
SHD<\/span><\/section>\n<\/td>\n
\n
\u901a\u8fc7\u5c06\u4e00\u4e2a\u56fe\u8f6c\u6362\u4e3a\u53e6\u4e00\u4e2a\u56fe\u6240\u9700\u7684\u53d8\u5316\u7684\u6570\u91cf\u6765\u6bd4\u8f83\u4e24\u4e2a\u56fe\u7684\u5ea6\u91cf\uff0c\u88ab\u5e7f\u6cdb\u4f7f\u7528\uff0c\u4f46\u504f\u5411\u4e8e\u8bc6\u522b\u8fb9\u7684\u654f\u611f\u6027\u800c\u4e0d\u662f\u7279\u5f02\u6027\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
\u5e73\u8861\u8ba1\u5206\u51fd\u6570<\/span><\/section>\n<\/td>\n
\n
\u901a\u8fc7\u8003\u8651\u6df7\u6dc6\u77e9\u9635\u7684\u6240\u6709\u5143\u7d20\u6765\u6bd4\u8f83\u4e24\u4e2a\u89e3\u51b3SHD\u4e2d\u504f\u5dee\u7684\u56fe\u7684\u5ea6\u91cf\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n
\n
\u5206\u7c7b\u6307\u6807<\/span><\/section>\n<\/td>\n
\n
\u53ef\u4ee5\u4f7f\u7528\u7279\u5b9a\u4e8e\u5206\u7c7b\u7684\u6307\u6807\u5982AUC\u6216TPR (\u89c1\u4e0a)\u6765\u6bd4\u8f83\u56fe\uff0c\u5373\u771f\u8fb9\u7387=\u771f\u9633\u6027\u7387\uff0c\u5047\u8fb9\u6570=\u5047\u9633\u6027\uff0c\u6f0f\u8fb9\u7387=\u5047\u9634\u6027\u3002<\/span><\/section>\n<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n
<\/span><\/section>\n
<\/span><\/section>\n
 <\/span><\/section>\n
\u867d\u7136\uff0c\u751f\u6210\u7684\u6a21\u578b\u53ef\u4ee5\u5728ML\u6846\u67b6\u5185\u8fdb\u884c\u8bc4\u4f30\uff0c\u5373\u6bd4\u8f83\u4e00\u7cfb\u5217\u6a21\u578b\u5e76\u4f7f\u7528\u6027\u80fd\u6307\u6807\u6765\u9009\u62e9\u6a21\u578b\u3002\u4f46\u76f8\u6bd4\u4e8e\u5982\u903b\u8f91\u56de\u5f52\u7684\u4f20\u7edf\u7edf\u8ba1\u65b9\u6cd5<\/span>\uff08\u5982\u903b\u8f91\u56de\u5f52\uff09<\/span>\u7684\u6a21\u578b\uff0c\u57fa\u4e8eML\u7684\u6a21\u578b\u4ecd\u7136\u662f\u76ee\u524d\u6280\u672f\u7684\u6700\u9ad8\u6c34\u5e73\u3002\u4e00\u4e9b\u6027\u80fd\u6307\u6807<\/span>\uff08\u5982\u4e0e\u5206\u7c7b\u6709\u5173\u7684\u654f\u611f\u6027\u548c\u7279\u5f02\u6027\uff09<\/span>\u9002\u7528\u4e8eML\u548c\u975eML\u65b9\u6cd5\uff0c\u5e76\u4e14\u53ef\u4ee5\u8de8\u6846\u67b6\u4f7f\u7528\u3002<\/span><\/section>\n

<\/span><\/section>\n


<\/p>\n

\n
\n
\n
\n
\n

<\/section>\n

<\/section>\n
\n
<\/svg><\/section>\n<\/section>\n<\/section>\n
\n

8. \u672a\u6765\u5c55\u671b<\/strong><\/strong><\/p>\n<\/section>\n

\n

<\/section>\n

<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/h3>\n

<\/section>\n
\u901a\u8fc7\u4f7f\u7528\u63cf\u8ff0\u3001\u9884\u6d4b\u548c\u56e0\u679c\u63a8\u65ad\u7b49\u4e0d\u540c\u7684\u7814\u7a76\u76ee\u7684\uff0c\u53ef\u4ee5\u5c06\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u4e2d\u7684\u5e38\u89c1\u7814\u7a76\u95ee\u9898\u6620\u5c04\u5230\u9002\u5f53\u7684ML\u65b9\u6cd5\u4e2d\u3002\u672c\u6587\u4e3a\u8fd9\u4e9b\u7814\u7a76\u76ee\u7684\u63d0\u4f9b\u4e86\u4e00\u4e2a\u65b0\u7684\u603b\u7ed3\uff0c\u5c06\u4f7fML\u5728\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u4e2d\u5f97\u5230\u66f4\u7cfb\u7edf\u7684\u4f7f\u7528\u3002\u867d\u7136\u672c\u6587\u662f\u8fc4\u4eca\u4e3a\u6b62\u5bf9ML\u5728\u793e\u4f1a\u548c\u5065\u5eb7\u79d1\u5b66\u4e2d\u5e94\u7528\u7684\u6700\u5168\u9762\u6982\u8ff0\uff0c\u4f46\u672c\u6587\u672a\u5305\u542b\u4f7f\u7528ML\u7684\u6311\u6218\uff0c\u4f8b\u5982\u7814\u7a76\u57fa\u7840\u8bbe\u65bd\u548c\u8ba1\u7b97\u8981\u6c42\uff0c\u4ee5\u53ca\u9690\u79c1\u548c\u6570\u636e\u4fdd\u62a4\u7684\u95ee\u9898\u3002\u672c\u6587\u4e5f\u53ea\u662f\u7565\u5fae\u89e6\u53ca\u4e86\u4e3a\u66f4\u81ea\u52a8\u3001\u66f4\u5feb\u901f\u5730\u5904\u7406\u5927\u6570\u636e\u800c\u5f00\u53d1\u7684ML\u7b97\u6cd5\uff0c\u4f8b\u5982\u8bed\u97f3\u6216\u6587\u672c\u8bc6\u522b\u7684\u81ea\u7136\u8bed\u8a00\u5904\u7406\u548c\u60c5\u611f\u5206\u6790\uff0c\u5e76\u53ea\u63d0\u4f9b\u4e86\u4e00\u4e9b\u4e0e\u5168\u57fa\u56e0\u6570\u636e\u5206\u6790\u6709\u5173\u7684\u4f8b\u5b50\u3002\u5c06SNA\u5e94\u7528\u4e8e\u65c5\u6e38\u6216\u8425\u9500\u5728\u5176\u4ed6\u8bba\u6587\u8ba8\u8bba\u8fc7\uff0853\uff09\u3002<\/span><\/section>\n

<\/span><\/section>\n
\u5982\u679c\u901a\u8fc7\u6df1\u601d\u719f\u8651\u5730\u5c06\u5176\u5e94\u7528\u4e8e\u76f8\u5173\u95ee\u9898\uff0cML\u65b9\u6cd5\u6709\u53ef\u80fd\u6781\u5927\u5730\u6539\u5584\u7ecf\u9a8c\u5206\u6790\uff0c\u6700\u597d\u662f\u901a\u8fc7\u5728\u793e\u4f1a\u6216\u5065\u5eb7\u79d1\u5b66\u9886\u57df\u53d7\u8fc7\u8bad\u7ec3\u7684\u7814\u7a76\u4eba\u5458\u4e0e\u53d7\u8fc7\u65b9\u6cd5\u5b66\u8bad\u7ec3\u7684\u7814\u7a76\u4eba\u5458\u4e4b\u95f4\u7684\u5408\u4f5c\u3002\u5e0c\u671b\u672c\u6587\u80fd\u591f\u7cfb\u7edf\u603b\u7ed3\u5e76\u63a8\u52a8\u6700\u8fd1\u5728\u793e\u4f1a\u548c\u5065\u5eb7\u7814\u7a76\u4e2d\u53d1\u5c55\u8d77\u6765\u7684ML\u65b9\u6cd5\u3002<\/span><\/section>\n

<\/span><\/section>\n

<\/section>\n
\n
\n
\n

\u53c2\u8003\u6587\u732e<\/strong><\/p>\n<\/section>\n<\/section>\n<\/section>\n

\n
\n
\n
\n
1. T. L. Wiemken, R. R. Kelley, Machine learning in epidemiology and health outcomes<\/span><\/section>\n
research. Ann. Rev. Public Health 41, 21\u201336 (2020).<\/span><\/section>\n
2. T. Yarkoni, J. Westfall, Choosing prediction over explanation in psychology: Lessons<\/span><\/section>\n
from machine learning. Perspect. Psychol. Sci. 12, 1100\u20131122 (2017).<\/span><\/section>\n
3. H. R. Varian, Big data: New tricks for econometrics. J. Econ. Perspect. 28, 3\u201328 (2014).<\/span><\/section>\n
4. J. Friedman, T. Hastie, R. Tibshirani, The Elements of Statistical Learning (Springer Series in Statistics, Springer, 2001), vol. 1.<\/span><\/section>\n
5. G. James, D. Witten, T. Hastie, R. Tibshirani, An Introduction to Statistical Learning<\/span><\/section>\n
(Springer, 2013), vol. 112.<\/span><\/section>\n
6. M. A. Hern\u00e1n, J. Hsu, B. Healy, A second chance to get causal inference right:<\/span><\/section>\n
A classification of data science tasks. Chance 32, 42\u201349 (2019).<\/span><\/section>\n
7. J. Tukey, Exploratory Data Analysis (Addison-Wesley Series in Behavioral Science:<\/span><\/section>\n
Quantitative Methods, Addison-Wesley, 1977).<\/span><\/section>\n
8. I. Y. Chen, E. Pierson, S. Rose, S. Joshi, K. Ferryman, M. Ghassemi, Ethical machine learning in healthcare. Annu. Rev. Biomed. Data Sci. 4, 123\u2013144 (2021).<\/span><\/section>\n
9. W. J. Murdoch, C. Singh, K. Kumbier, R. Abbasi-Asl, B. Yu, Definitions, methods,<\/span><\/section>\n
and applications in interpretable machine learning. Proc. Natl. Acad. Sci. U.S.A. 116,<\/span><\/section>\n
22071\u201322080 (2019).<\/span><\/section>\n
10. S. J. Pan, Q. Yang, A survey on transfer learning. IEEE Trans. Knowl. Data Eng. 22,<\/span><\/section>\n
1345\u20131359 (2009).<\/span><\/section>\n
11. Y. Gao, Y. Cui, Deep transfer learning for reducing health care disparities arising<\/span><\/section>\n
from biomedical data inequality. Nat. Commun. 11, 5131 (2020).<\/span><\/section>\n
12. M. Kuhn, K. Johnson, Applied Predictive Modeling (Springer, 2013), vol. 26.<\/span><\/section>\n
13. D. Donoho, J. Tanner, Observed universality of phase transitions in high-dimensional<\/span><\/section>\n
geometry, with implications for modern data analysis and signal processing. Phil. Trans.<\/span><\/section>\n
R. Soc. A 367, 4273\u20134293 (2009).<\/span><\/section>\n
14. C. Ding, H. Peng, Minimum redundancy feature selection from microarray gene expression data. J. Bioinform. Comput. Biol. 3, 185\u2013205 (2005).<\/span><\/section>\n
15. F. Yu, C. Wei, P. Deng, T. Peng, X. Hu, Deep exploration of random forest model boosts the interpretability of machine learning studies of complicated immune responses and lung burden of nanoparticles. Sci. Adv. 7, eabf4130 (2021)<\/span><\/section>\n
16. M. Anthony, P. Bartlett, Neural Network Learning: Theoretical Foundations (Cambridge Univ. Press, 1999).<\/span><\/section>\n
17. M. A. Hern\u00e1n, Does water kill? A call for less casual causal inferences. Ann. Epidemiol. 26, 674\u2013680 (2016).<\/span><\/section>\n
18. R. D. King, O. I. Orhobor, C. C. Taylor, Cross-validation is safe to use. Nat. Mach. Intell. 3, 276 (2021).<\/span><\/section>\n
19. E. LeDell, S. Poirier, H2O AutoML: Scalable automatic machine learning. 7th ICML<\/span><\/section>\n
Workshop on Automated Machine Learning (AutoML); <\/span>www.automl.org\/wp-content\/<\/section>\n
uploads\/2020\/07\/AutoML_2020_paper_61.pdf.<\/span><\/section>\n
20. F. Pedregosa, G. Vaorquaux, A. Gramfort, V. Michel, B. Thirion, O. Grisel, M. Blondel,<\/span><\/section>\n
P. Prettenhofer, R. Weiss, V. Dubourg, J. Vanderplas, Scikit-learn: Machine learning<\/span><\/section>\n
in Python. J. Mach. Learn. Res. 12, 2825\u20132830 (2011).<\/span><\/section>\n
21. European Commission, Directorate General for Communications Networks, Content and Technology, High Level Expert Group on Artificial Intelligence, Ethics guidelines for<\/span><\/section>\n
trustworthy AI (Publications Office, LU, 2019); <\/span>https:\/\/data.europa.eu\/doi\/10.2759\/177365.<\/section>\n
22. C. Rudin, Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead. Nat. Mach. Intell. 1, 206\u2013215 (2019).<\/span><\/section>\n
23. K. J. Rohlfing, P. Cimiano, I. Scharlau, T. Matzner, H. M. Buhl, H. Buschmeier, E. Esposito, A. Grimminger, B. Hammer, R. H\u00e4b-Umbach, Explanation as a social practice: Toward a conceptual framework for the social design of AI systems. IEEE Trans. Cogn. Dev. Syst. 13, 717\u2013728 (2020).<\/span><\/section>\n
24. B. M. Greenwell, pdp: An R package for constructing partial dependence plots. R J. 9, 421\u2013436 (2017).<\/span><\/section>\n
25. S. M. Lundberg, G. G. Erion, S.-I. Lee, Consistent individualized feature attribution for tree ensembles. arXiv:1802.03888 [cs.LG] (12 February 2018).<\/span><\/section>\n
26. S. Mangalathu, S.-H. Hwang, J.-S. Jeon, Failure mode and effects analysis of RC members based on machine-learning-based SHapley additive exPlanations (SHAP) approach. Eng. Struct. 219, 110927 (2020).<\/span><\/section>\n
27. P. Hall, N. Gill, M. Kurka, W. Phan, Machine learning interpretability with h2o driverless ai (H2O.ai, 2017).<\/span><\/section>\n
28. S. Barocas, M. Hardt, A. Narayanan, Fairness in Machine Learning (fairmlbook.org, 2019), vol. 1.<\/span><\/section>\n
29. Z. Obermeyer, B. Powers, C. Vogeli, S. Mullainathan, Dissecting racial bias in an algorithm used to manage the health of populations. Science 366, 447\u2013453 (2019).<\/span><\/section>\n
30. S. Mitchell, E. Potash, S. Barocas, A. D\u2019Amour, K. Lum, Algorithmic fairness: Choices,<\/span><\/section>\n
assumptions, and definitions. Annu. Rev. Stat. Appl. 8, 141\u2013163 (2021).<\/span><\/section>\n
31. S. R. Pfohl, A. Foryciarz, N. H. Shah, An empirical characterization of fair machine learning for clinical risk prediction. J. Biomed. Inform. 113, 103621 (2021).<\/span><\/section>\n
32. C. Birkenbihl, M. A. Emon, H. Vrooman, S. Westwood, S. Lovestone; AddNeuroMed<\/span><\/section>\n
Consortium, M. Hofmann-Apitius, H. Fr\u00f6hlich, Differences in cohort study data affect<\/span><\/section>\n
external validation of artificial intelligence models for predictive diagnostics of<\/span><\/section>\n
dementia-lessons for translation into clinical practice. EPMA J. 11, 367\u2013376 (2020).<\/span><\/section>\n
33. A. Tsymbal, The problem of concept drift: Definitions and related work. Computer Science Department, Trinity College Dublin. 106, 58 (2004).<\/span><\/section>\n
34. M. Straat, F. Abadi, Z. Kan, C. G\u00f6pfert, B. Hammer, M. Biehl, Supervised learning in the presence of concept drift: A modelling framework. Neural Comput. Appl. 34, 101\u2013118 (2022).<\/span><\/section>\n
35. S. Salmi, S. M\u00e9relle, R. Gilissen, W.-P. Brinkman, Content-based recommender support system for counselors in a suicide prevention chat helpline: Design and evaluation study. J. Med. Internet Res. 23, e21690 (2021).<\/span><\/section>\n
36. L. Van Der Maaten, E. Postma, J. Van den Herik, Dimensionality reduction: A comparative. J. Mach. Learn. Res. 10, 66\u201371 (2009).<\/span><\/section>\n
37. V. Kuan, H. C. Fraser, M. Hingorani, S. Denaxas, A. Gonzalez-Izquierdo, K. Direk, D. Nitsch, R. Mathur, C. A. Parisinos, R. T. Lumbers, Data-driven identification of ageing-related diseases from electronic health records. Sci. Rep. 11, 1\u201317 (2021).<\/span><\/section>\n
38. S. M. Lundberg, B. Nair, M. S. Vavilala, M. Horibe, M. J. Eisses, T. Adams, D. E. Liston,<\/span><\/section>\n
D. K.-W. Low, S.-F. Newman, J. Kim, Explainable machine-learning predictions for the<\/span><\/section>\n
prevention of hypoxaemia during surgery. Nat. Biomed. Eng. 2, 749\u2013760 (2018).<\/span><\/section>\n
39. N. Demnitz, M. Anat\u00fcrk, C. L. Allan, N. Filippini, L. Griffanti, C. E. Mackay, A. Mahmood, C. E. Sexton, S. Suri, A. G. Topiwala, E. Zsoldos, M. Kivim\u00e4ki, A. Singh-Manoux, K. P. Ebmeier, Association of trajectories of depressive symptoms with vascular risk, cognitive function and adverse brain outcomes: The Whitehall II MRI sub-study.<\/span><\/section>\n
J. Psychiatr. Res. 131, 85\u201393 (2020).<\/span><\/section>\n
40. S. Amoretti, N. Verdolini, G. Mezquida, F. D. Rabelo-da-Ponte, M. J. Cuesta, L. Pina-Camacho, M. Gomez-Ramiro, C. De-la-C\u00e1mara, A. Gonz\u00e1lez-Pinto, C. M. D\u00edaz-Caneja, I. Corripio, E. Vieta, E. de la Serna, A. Man\u00e9, B. Sol\u00e9, A. F. Carvalho, M. Serra, M. Bernardo,<\/span><\/section>\n
Identifying clinical clusters with distinct trajectories in first-episode psychosis through<\/span><\/section>\n
an unsupervised machine learning technique. Eur. Neuropsychopharmacol. 47, 112\u2013129<\/span><\/section>\n
(2021).<\/span><\/section>\n
41. J. Weiss, E. Puterman, A. A. Prather, E. B. Ware, D. H. Rehkopf, A data-driven prospective study of dementia among older adults in the United States. PLOS ONE 15, e0239994 (2020).<\/span><\/section>\n
42. M. Hern\u00e1n, J. Robins, Causal Inference: What If (Chapman & Hall\/CRC, 2020).<\/span><\/section>\n
43. A. Galozy, Towards understanding ICU procedures using similarities in patient<\/span><\/section>\n
trajectories: An exploratory study on the MIMIC-III intensive care database (2018);<\/span><\/section>\n
www.diva-portal.org\/smash\/get\/diva2:1229433\/FULLTEXT02.<\/section>\n
44. W. Luo, T. Nguyen, M. Nichols, T. Tran, S. Rana, S. Gupta, D. Phung, S. Venkatesh,<\/span><\/section>\n
S. Allender, Is demography destiny? Application of machine learning techniques<\/span><\/section>\n
to accurately predict population health outcomes from a minimal demographic dataset.<\/span><\/section>\n
PLOS ONE 10, e0125602 (2015).<\/span><\/section>\n
45. H. Wang, A. Rodr\u00edguez, Identifying pediatric cancer clusters in florida using log-linear models and generalized lasso penalties. Stat. Public Policy 1, 86\u201396 (2014).<\/span><\/section>\n
46. E. L. Aiken, A. T. Nguyen, C. Viboud, M. Santillana, Toward the use of neural networks<\/span><\/section>\n
for influenza prediction at multiple spatial resolutions. Sci. Adv. 7, eabb1237 (2021).<\/span><\/section>\n
47. J. C. Eichstaedt, R. J. Smith, R. M. Merchant, L. H. Ungar, P. Crutchley, D. Preo\u0163iuc-Pietro, D. A. Asch, H. A. Schwartz, Facebook language predicts depression in medical records. Proc. Natl. Acad. Sci. U.S.A. 115, 11203\u201311208 (2018).<\/span><\/section>\n
48. L. C. de Langavant, E. Bayen, A. Bachoud-L\u00e9vi, K. Yaffe, Approximating dementia<\/span><\/section>\n
prevalence in population-based surveys of aging worldwide: An unsupervised machine<\/span><\/section>\n
learning approach. Alzheimers Dementia 6, e12074 (2020).<\/span><\/section>\n
49. S. I. R. Bhagyashree, K. Nagaraj, M. Prince, C. H. Fall, M. Krishna, Diagnosis of Dementia by Machine learning methods in epidemiological studies: A pilot exploratory study from south India. Soc. Psychiatry Psychiatr. Epidemiol. 53, 77\u201386 (2018).<\/span><\/section>\n
50. K. Z. Gianattasio, A. Ciarleglio, M. C. Power, Development of algorithmic dementia<\/span><\/section>\n
ascertainment for racial\/ethnic disparities research in the US health and retirement study. Epidemiology 31, 126\u2013133 (2020).<\/span><\/section>\n
51. M. Morris, M. S. Handcock, D. R. Hunter, Specification of exponential-family random<\/span><\/section>\n
graph models: Terms and computational aspects. J. Stat. Softw. 24, 1548\u20137660 (2008).<\/span><\/section>\n
52. P. Barber\u00e1, J. T. Jost, J. Nagler, J. A. Tucker, R. Bonneau, Tweeting from left to right: Is<\/span><\/section>\n
online political communication more than an echo chamber? Psychol. Sci. 26, 1531\u20131542 (2015).<\/span><\/section>\n
53. D. Camacho, A. Panizo-LLedot, G. Bello-Orgaz, A. Gonzalez-Pardo, E. Cambria, The four dimensions of social network analysis: An overview of research methods, applications, and software tools. Inf. Fusion 63, 88\u2013120 (2020).<\/span><\/section>\n
54. J. Fan, R. Li, Variable selection via nonconcave penalized likelihood and its oracle<\/span><\/section>\n
properties. J. Am. Stat. Assoc. 96, 1348\u20131360 (2001).<\/span><\/section>\n
55. J. L. Hill, A. Linero, J. Murray, Bayesian additive regression trees: A review and look forward. Annu. Rev. Stat. Appl. 7, 251\u2013278 (2020).<\/span><\/section>\n
56. V. Dorie, M. Harada, N. B. Carnegie, J. L. Hill, A flexible, interpretable framework<\/span><\/section>\n
for assessing sensitivity to unmeasured confounding. Stat. Med. 35, 3453\u20133470 (2016).<\/span><\/section>\n
57. V. Svetnik, A. Liaw, C. Tong, J. C. Culberson, R. P. Sheridan, B. P. Feuston, Random forest: A classification and regression tool for compound classification and QSAR modeling. J. Chem. Inf. Comput. Sci. 43, 1947\u20131958 (2003).<\/span><\/section>\n
58. B. Seligman, S. Tuljapurkar, D. Rehkopf, Machine learning approaches to the social<\/span><\/section>\n
determinants of health in the health and retirement study. SSM Popul. Health 4, 95\u201399 (2018).<\/span><\/section>\n
59. J. H. Friedman, Stochastic gradient boosting. Comput. Stat. Data Anal. 38, 367\u2013378 (2002).<\/span><\/section>\n
60. T. Hothorn, P. B\u00fchlmann, T. Kneib, M. Schmid, B. Hofner, mboost: Model-based boosting (R package version, 2012).<\/span><\/section>\n
61. A. I. Naimi, L. B. Balzer, Stacked generalization: An introduction to super learning.<\/span><\/section>\n
Eur. J. Epidemiol. 33, 459\u2013464 (2018).<\/span><\/section>\n
62. H. S. Laqueur, A. B. Shev, R. M. C. Kagawa, SuperMICE: An ensemble machine learning approach to multiple imputation by chained equations. Am. J. Epidemiol. 191, 516\u2013525 (2022).<\/span><\/section>\n
63. T. Fine, M. Hassoun, Fundamentals of artificial neural networks [Book Reviews]. IEEE Trans. Inf. Theory 42, 1322\u20131324 (1996).<\/span><\/section>\n
64. S. Dreiseitl, L. Ohno-Machado, Logistic regression and artificial neural network<\/span><\/section>\n
classification models: A methodology review. J. Biomed. Inform. 35, 352\u2013359 (2002).<\/span><\/section>\n
65. F. Harrell, Regression Modeling Strategies. With Applications to Linear Models, Logistic and Ordinal Regression, and Survival Analysis (Springer, 2015).<\/span><\/section>\n
66. R. Neal, Bayesian Learning for Neural Networks (Springer, 2012).<\/span><\/section>\n
67. C. Szegedy, W. Zaremba, I. Sutskever, J. Bruna, D. Erhan, I. Goodfellow, R. Fergus,<\/span><\/section>\n
Intriguing properties of neural networks. arXiv:1312.6199 [cs.CV] (19 February 2014).<\/span><\/section>\n
68. G. Carbone, M. Wicker, L. Laurenti, A. Patane, L. Bortolussi, G. Sanguinetti, Robustness of bayesian neural networks to gradient-based attacks. Adv. Neural Inf. Process. Syst. 33, 15602\u201315613 (2020).<\/span><\/section>\n
69. I. Kavakiotis, O. Tsave, A. Salifoglou, N. Maglaveras, I. Vlahavas, I. Chouvarda, Machine learning and data mining methods in diabetes research. Comput. Struct. Biotechnol. J. 15, 104\u2013116 (2017).<\/span><\/section>\n
70. M. Beeksma, S. Verberne, A. van den Bosch, E. Das, I. Hendrickx, S. Groenewoud,<\/span><\/section>\n
Predicting life expectancy with a long short-term memory recurrent neural network<\/span><\/section>\n
using electronic medical records. BMC Med. Inform. Decis. Mak. 19, 36 (2019).<\/span><\/section>\n
71. R. C. Kessler, I. Hwang, C. A. Hoffmire, J. F. McCarthy, M. V. Petukhova, A. J. Rosellini,<\/span><\/section>\n
N. A. Sampson, A. L. Schneider, P. A. Bradley, I. R. Katz, C. Thompson, R. M. Bossarte,<\/span><\/section>\n
Developing a practical suicide risk prediction model for targeting high-risk patients<\/span><\/section>\n
in the Veterans Health Administration. Int. J. Methods Psychiatr. Res. 26, e1575 (2017).<\/span><\/section>\n
72. S. Rose, Mortality risk score prediction in an elderly population using machine learning. Am. J. Epidemiol. 177, 443\u2013452 (2013).<\/span><\/section>\n
73. E. Christodoulou, J. Ma, G. S. Collins, E. W. Steyerberg, J. Y. Verbakel, B. Van Calster,<\/span><\/section>\n
A systematic review shows no performance benefit of machine learning over logistic<\/span><\/section>\n
regression for clinical prediction models. J. Clin. Epidemiol. 110, 12\u201322 (2019).<\/span><\/section>\n
74. V. Ba\u0107ak, E. H. Kennedy, Principled machine learning using the super learner:<\/span><\/section>\n
An application to predicting prison violence. Sociol. Methods Res. 48, 698\u2013721 (2019).<\/span><\/section>\n
75. J. Blumenstock, G. Cadamuro, R. On, Predicting poverty and wealth from mobile phone metadata. Science 350, 1073\u20131076 (2015).<\/span><\/section>\n
76. A. M. Alaa, T. Bolton, E. Di Angelantonio, J. H. Rudd, M. van der Schaar, Cardiovascular disease risk prediction using automated machine learning: A prospective study of 423,604 UK Biobank participants. PLOS ONE 14, e0213653 (2019).<\/span><\/section>\n
77. S. Licher, M. J. Leening, P. Yilmaz, F. J. Wolters, J. Heeringa, P. J. Bindels; Alzheimer\u2019s Disease Neuroimaging Initiative, M. W. Vernooij, B. C. Stephan, E. W. Steyerberg, M. Kamran Ikram, M. Arfan Ikram, Development and validation of a dementia risk prediction model in the general population: An analysis of three longitudinal studies. Am. J. Psychiatry 176, 543\u2013551 (2019).<\/span><\/section>\n
78. D. E. Goin, K. E. Rudolph, J. Ahern, Predictors of firearm violence in urban communities: A machine-learning approach. Health Place 51, 61\u201367 (2018).<\/span><\/section>\n
79. N. V. Chawla, K. W. Bowyer, L. O. Hall, W. P. Kegelmeyer, SMOTE: Synthetic minority<\/span><\/section>\n
over-sampling technique. J. Artif. Intell. Res. 16, 321\u2013357 (2002).<\/span><\/section>\n
80. H. Kvamme, N. Sellereite, K. Aas, S. Sjursen, Predicting mortgage default using<\/span><\/section>\n
convolutional neural networks. Expert Syst. Appl. 102, 207\u2013217 (2018).<\/span><\/section>\n
81. M. Khalilia, S. Chakraborty, M. Popescu, Predicting disease risks from highly imbalanced data using random forest. BMC Med. Inform. Decis. Mak. 11, 51 (2011).<\/span><\/section>\n
82. D. Facal, S. Valladares-Rodriguez, C. Lojo-Seoane, A. X. Pereiro, L. Anido-Rifon,<\/span><\/section>\n
O. Juncos-Rabad\u00e1n, Machine learning approaches to studying the role of cognitive<\/span><\/section>\n
reserve in conversion from mild cognitive impairment to dementia. Int. J. Geriatr. Psychiatry 34, 941\u2013949 (2019).<\/span><\/section>\n
83. M. Grassi, D. A. Loewenstein, D. Caldirola, K. Schruers, R. Duara, G. Perna, A<\/span><\/section>\n
clinically-translatable machine learning algorithm for the prediction of Alzheimer\u2019s<\/span><\/section>\n
disease conversion: Further evidence of its accuracy via a transfer learning approach.<\/span><\/section>\n
Int. Psychogeriatr. 31, 937\u2013945 (2019).<\/span><\/section>\n
84. R. V. Marinescu, N. P. Oxtoby, A. L. Young, E. E. Bron, A. W. Toga, M. W. Weiner, F. Barkhof, N. C. Fox, A. Eshaghi, T. Toni, M. Salaterski, V. Lunina, M. Ansart, S. Durrleman, P. Lu, S. Iddi, D. Li, W. K. Thompson, M. C. Donohue, A. Nahon, Y. Levy, D. Halbersberg, M. Cohen, H. Liao, T. Li, K. Yu, H. Zhu, J. G. Tamez-Pena, A. Ismail, T. Wood, H. C. Bravo, M. Nguyen,N. Sun, J. Feng, B. T. Thomas Yeo, G. Chen, K. Qi, S. Chen, D. Qiu, I. Buciuman, A. Kelner, R. Pop, D. Rimocea, M. M. Ghazi, M. Nielsen, S. Ourselin, L. Sorensen, V. Venkatraghavan, K. Liu, C. Rabe, P. Manser, S. M. Hill, J. Howlett, Z. Huang, S. Kiddle, S. Mukherjee, A. Rouanet, B. Taschler, B. D. M. Tom, S. R. White, N. Faux, S. Sedai, Javier de Velasco Oriol, E. E. V. Clemente, K. Estrada, L. Aksman, A. Altmann, C. M. Stonnington, Y. Wang, J. Wu, V. Devadas, C. Fourrier, L. L. Raket, A. Sotiras, G. Erus, J. Doshi, C. Davatzikos, J. Vogel, A. Doyle, A. Tam, A. Diaz-Papkovich, E. Jammeh, I. Koval, P. Moore, T. J. Lyons, J. Gallacher, J. Tohka, R. Ciszek, B. Jedynak, K. Pandya, M. Bilgel, W. Engels, J. Cole, P. Golland, S. Klein, D. C. Alexander, The Alzheimer\u2019s disease prediction of longitudinal evolution (TADPOLE) challenge: Results after 1 year follow-up. arXiv:2002.03419 [q-bio.PE] (9 February 2020).<\/span><\/section>\n
85. K. Z. LeWinn, N. R. Bush, A. Batra, F. Tylavsky, D. Rehkopf, Identification of modifiable<\/span><\/section>\n
social and behavioral factors associated with childhood cognitive performance.<\/span><\/section>\n
JAMA Pediatr. 174, 1063\u20131072 (2020).<\/span><\/section>\n
86. O. Weller, L. Sagers, C. Hanson, M. Barnes, Q. Snell, E. S. Tass, Predicting suicidal thoughts and behavior among adolescents using the risk and protective factor framework: A large-scale machine learning approach. PLOS ONE 16, e0258535 (2021).<\/span><\/section>\n
87. N. Kimura, Y. Aso, K. Yabuuchi, M. Ishibashi, D. Hori, Y. Sasaki, A. Nakamichi, S. Uesugi, H. Fujioka, S. Iwao, M. Jikumaru, T. Katayama, K. Sumi, A. Eguchi, S. Nonaka, M. Kakumu, E. Matsubara, Modifiable lifestyle factors and cognitive function in older people: A cross-sectional observational study. Front. Neurol. 10, 401 (2019).<\/span><\/section>\n
88. R. C. Schell, B. Allen, W. C. Goedel, B. D. Hallowell, R. Scagos, Y. Li, M. S. Krieger, D. B. Neill, B. D. Marshall, M. Cerda, J. Ahern, Identifying predictors of opioid overdose death at a neighborhood level with machine learning. Am. J. Epidemiol. 191, 526\u2013533 (2022).<\/span><\/section>\n
89. J. M. Platt, K. A. McLaughlin, A. R. Luedtke, J. Ahern, A. S. Kaufman, K. M. Keyes, Targeted estimation of the relationship between childhood adversity and fluid intelligence in a US population sample of adolescents. Am. J. Epidemiol. 187, 1456\u20131466 (2018).<\/span><\/section>\n
90. J. C. Peterson, D. D. Bourgin, M. Agrawal, D. Reichman, T. L. Griffiths, Using large-scale experiments and machine learning to discover theories of human decision-making. Science 372, 1209\u20131214 (2021).<\/span><\/section>\n
91. T. Ploner, S. He\u00df, M. Grum, P. Drewe-Boss, J. Walker, Using gradient boosting with stability selection on health insurance claims data to identify disease trajectories in chronic obstructive pulmonary disease. Stat. Methods Med. Res. 29, 3684\u20133694 (2020).<\/span><\/section>\n
92. M. Anat\u00fcrk, T. Kaufmann, J. H. Cole, S. Suri, L. Griffanti, E. Zsoldos, N. Filippini,<\/span><\/section>\n
A. Singh-Manoux, M. Kivim\u00e4ki, L. T. Westlye, K. P. Ebmeier, A.-M. G. de Lange, Prediction<\/span><\/section>\n
of brain age and cognitive age: Quantifying brain and cognitive maintenance in aging.<\/span><\/section>\n
Hum. Brain Mapp. 42, 1626\u20131640 (2020).<\/span><\/section>\n
93. M. M. Glymour, Using causal diagrams to understand common problems in social<\/span><\/section>\n
epidemiology, in Methods in Social Epidemiology, J. M. Oakes, J. S. Kaufman, Eds.<\/span><\/section>\n
(Wiley, ed. 2, 2006) pp. 393\u2013428.<\/span><\/section>\n
94. E. C. Matthay, M. M. Glymour, A graphical catalog of threats to validity: Linking social science with epidemiology. Epidemiology 31, 376\u2013384 (2020).<\/span><\/section>\n
95. P. W. Tennant, W. J. Harrison, E. J. Murray, K. F. Arnold, L. Berrie, M. P. Fox, S. C. Gadd,<\/span><\/section>\n
C. Keeble, L. R. Ranker, J. Textor, Use of directed acyclic graphs (DAGs) in applied<\/span><\/section>\n
health research: Review and recommendations. medRxiv 2019.12.20.19015511 (2019).<\/span><\/section>\n
96. J. M. Robins, M. \u00c1. Hern\u00e1n, B. Brumback, Marginal structural models and causal inference in epidemiology. Epidemiology 11, 550\u2013560 (2000).<\/span><\/section>\n
97. X. Garc\u00eda-Alb\u00e9niz, J. Hsu, M. A. Hern\u00e1n, The value of explicitly emulating a target trial when using real world evidence: An application to colorectal cancer screening. Eur. J. Epidemiol. 32, 495\u2013500 (2017).<\/span><\/section>\n
98. M. A. Hern\u00e1n, J. M. Robins, Using big data to emulate a target trial when a randomized trial is not available. Am. J. Epidemiol. 183, 758\u2013764 (2016).<\/span><\/section>\n
99. I. M. Aris, A. L. Sarvet, M. J. Stensrud, R. Neugebauer, L.-J. Li, M.-F. Hivert, E. Oken,<\/span><\/section>\n
J. G. Young, Separating algorithms from questions and causal inference with unmeasured exposures: An application to birth cohort studies of early body mass index rebound. Am. J. Epidemiol. 190, 1414\u20131423 (2021).<\/span><\/section>\n
100. G. Shmueli, To explain or to predict? Stat. Sci. 25, 289\u2013310 (2010).<\/span><\/section>\n
101. L. B. Balzer, M. L. Petersen, Machine learning in causal inference: How do I love thee? Let me count the ways. Am. J. Epidemiol. 190, 1483\u20131487 (2021).<\/span><\/section>\n
102. T. Blakely, J. Lynch, K. Simons, R. Bentley, S. Rose, Reflection on modern methods: When worlds collide\u2014Prediction, machine learning and causal inference. Int. J. Epidemiol. 49, 2058\u20132064 (2020).<\/span><\/section>\n
103. H. A. Chipman, E. I. George, R. E. McCulloch, BART: Bayesian additive regression trees. Ann. Appl. Stat. 4, 266\u2013298 (2010).<\/span><\/section>\n
104. M. S. Schuler, S. Rose, Targeted maximum likelihood estimation for causal inference in observational studies. Am. J. Epidemiol. 185, 65\u201373 (2017).<\/span><\/section>\n
105. S. Wager, S. Athey, Estimation and inference of heterogeneous treatment effects using random forests. J. Am. Stat. Assoc. 113, 1228\u20131242 (2018).<\/span><\/section>\n
106. V. Chernozhukov, D. Chetverikov, M. Demirer, E. Duflo, C. Hansen, W. Newey, J. Robins, Double\/debiased machine learning for treatment and structural parameters. Econom. J. 21, C1\u2013C68 (2018).<\/span><\/section>\n
107. N. B. Carnegie, J. Wu, Variable selection and parameter tuning for BART modeling in the fragile families challenge. Socius 5, 2378023119825886 (2019).<\/span><\/section>\n
108. L. M. Bodnar, A. R. Cartus, S. I. Kirkpatrick, K. P. Himes, E. H. Kennedy, H. N. Simhan,<\/span><\/section>\n
W. A. Grobman, J. Y. Duffy, R. M. Silver, S. Parry, A. I. Naimi, Machine learning as a strategy to account for dietary synergy: An illustration based on dietary intake and adverse pregnancy outcomes. Am. J. Clin. Nutr. 111, 1235\u20131243 (2020).<\/span><\/section>\n
109. L. M. Bodnar, A. R. Cartus, E. H. Kennedy, S. I. Kirkpatrick, S. M. Parisi, K. P. Himes, C. B. Parker, W. A. Grobman, H. N. Simhan, R. M. Silver, D. A. Wing, S. Perry, A. I. Naimi, A doubly robust machine learning-based approach to evaluate body mass index as a modifier of the association between fruit and vegetable intake and preeclampsia. Am. J. Epidemiol. 191, 1396\u20131406 (2022).<\/span><\/section>\n
110. A. Decruyenaere, J. Steen, K. Colpaert, D. D. Benoit, J. Decruyenaere, S. Vansteelandt, The obesity paradox in critically ill patients: A causal learning approach to a casual finding. Crit. Care 24, 485 (2020).<\/span><\/section>\n
111. P. Brunori, G. Neidh\u00f6fer, The evolution of inequality of opportunity in Germany: A machine learning approach. Rev. Income Wealth 67, 900\u2013927 (2021).<\/span><\/section>\n
112. S. Rose, Intersections of machine learning and epidemiological methods for health<\/span><\/section>\n
services research. Int. J. Epidemiol. 49, 1763\u20131770 (2020).<\/span><\/section>\n
113. S. Athey, G. W. Imbens, Machine learning methods that economists should know about. Annu. Rev. Econom. 11, 685\u2013725 (2019).<\/span><\/section>\n
114. J. Delgadillo, P. Gonzalez, S. Duhne, Targeted prescription of cognitive\u2013behavioral<\/span><\/section>\n
therapy versus person-centered counseling for depression using a machine learning<\/span><\/section>\n
approach. J. Consult. Clin. Psychol. 88, 14\u201324 (2020).<\/span><\/section>\n
115. K. Imai, M. Ratkovic, Estimating treatment effect heterogeneity in randomized program evaluation. Ann. Appl. Stat. 7, 443\u2013470 (2013).<\/span><\/section>\n
116. U. Shalit, Can we learn individual-level treatment policies from clinical data? Biostatistics 21, 359\u2013362 (2020).<\/span><\/section>\n
117. J. L. Hill, Y.-S. Su, Assessing lack of common support in causal inference using Bayesian nonparametrics: Implications for evaluating the effect of breastfeeding on children\u2019s cognitive outcomes. Ann. Appl. Stat. 7, 1386\u20131420 (2013).<\/span><\/section>\n
118. L. Keele, D. S. Small, Comparing covariate prioritization via matching to machine<\/span><\/section>\n
learning methods for causal inference using five empirical applications. Am. Stat. 75,<\/span><\/section>\n
1\u20139 (2021).<\/span><\/section>\n
119. H. Ishwaran, U. B. Kogalur, E. H. Blackstone, M. S. Lauer, Random survival forests. Ann. Appl. Stat. 2, 841\u2013860 (2008).<\/span><\/section>\n
120. J. Pearl, Causality (Cambridge Univ. Press, 2009).<\/span><\/section>\n
121. M. M. Glymour, D. Spiegelman, Evaluating public health interventions: 5. Causal inference in public health research\u2014Do sex, race, and biological factors cause health outcomes? Am. J. Public Health 107, 81\u201385 (2017).<\/span><\/section>\n
122. R. Quintana, What race and gender stand for: Using Markov blankets to identify<\/span><\/section>\n
constitutive and mediating relationships. J. Comput. Soc. Sci. 5, 751\u2013779 (2021).<\/span><\/section>\n
123. R. M. Andrews, R. Foraita, V. Didelez, J. Witte, A practical guide to causal discovery with cohort data. arXiv:2108.13395 [stat.AP] (30 August 2021).<\/span><\/section>\n
124. J. Friedman, T. Hastie, R. Tibshirani, Sparse inverse covariance estimation with the<\/span><\/section>\n
graphical lasso. Biostatistics 9, 432\u2013441 (2008).<\/span><\/section>\n
125. Z. R. Li, T. H. McComick, S. J. Clark, Using Bayesian latent Gaussian graphical models to infer symptom associations in verbal autopsies. Bayesian Anal. 15, 781\u2013807 (2020).<\/span><\/section>\n
126. M. Scanagatta, A. Salmer\u00f3n, F. Stella, A survey on Bayesian network structure learning from data. Prog. Artif. Intell. 8, 425\u2013439 (2019).<\/span><\/section>\n
127. C. Heinze-Deml, M. H. Maathuis, N. Meinshausen, Causal structure learning. Annu. Rev. Stat Appl. 5, 371\u2013391 (2018).<\/span><\/section>\n
128. A. C. Constantinou, Y. Liu, K. Chobtham, Z. Guo, N. K. Kitson, Large-scale empirical validation of Bayesian network structure learning algorithms with noisy data. Int. J. Approx. Reason. 131, 151\u2013188 (2021).<\/span><\/section>\n
129. M. Scutari, J.-B. Denis, Bayesian Networks: With Examples in R (Chapman and Hall\/CRC, ed. 2, 2021).<\/span><\/section>\n
130. Z. Baranczuk, J. Estill, S. Blough, S. Meier, A. Merzouki, M. H. Maathuis, O. Keiser,<\/span><\/section>\n
Socio-behavioural characteristics and HIV: Findings from a graphical modelling analysis<\/span><\/section>\n
of 29 sub-Saharan African countries. J. Int. AIDS Soc. 22, e25437 (2019).<\/span><\/section>\n
131. M. Ester, H. Kriegel, J. Sander, X. Xu, A density-based algorithm for discovering clusters in large spatial databases with noise, in Proceedings of the Second International Conference on Knowledge Discovery and Data Mining (AAAI, 1996), vol. 240, p. 6.<\/span><\/section>\n
132. E. Schubert, J. Sander, M. Ester, H. P. Kriegel, X. Xu, DBSCAN revisited, revisited: Why and how you should (still) use DBSCAN. ACM Trans. Database Syst. 42, 1\u201321 (2017).<\/span><\/section>\n
133. M. Jordan, R. Jacobs, Hierarchical mixtures of experts and the EM algorithm. Neural Comput. 6, 181\u2013214 (1994).<\/span><\/section>\n
134. S. Masoudnia, R. Ebrahimpour, Mixture of experts: A literature survey. Artif. Intell. Rev. 42, 275\u2013293 (2014).<\/span><\/section>\n
135. M. E. Tipping, C. M. Bishop, Probabilistic principal component analysis. J. R. Stat. Soc. Series B 61, 611\u2013622 (1999).<\/span><\/section>\n
136. I. Goodfellow, J. Pouget-Abadie, M. Mirza, B. Xu, D. Warde-Farley, S. Ozair, A. Courville, Y. Bengio, Generative adversarial nets, in Advances in Neural Information Processing Systems, Z. Ghahramani, M. Welling, C. Cortes, N. Lawrence, K. Q. Weinberger, Eds. (Curran Associates Inc., 2014), vol. 27.<\/span><\/section>\n
137. D. P. Kingma, M. Welling, Auto-encoding variational bayes. arXiv:1312.6114 [stat.ML] (1 May 2014).<\/span><\/section>\n
138. V. A. Padilha, R. J. G. B. Campello, A systematic comparative evaluation of biclustering techniques. BMC Bioinform. 18, 55 (2017).<\/span><\/section>\n
139. G. E. Moran, V. Ro\u010dkov\u00e1, E. I. George, Spike-and-slab Lasso biclustering. Ann. Appl. Stat. 15, 148\u2013173 (2021).<\/span><\/section>\n
140. C. Cortes, V. Vapnik, Support-vector networks. Mach. Learn. 20, 273\u2013297 (1995).<\/span><\/section>\n
141. C. S. Bojer, J. P. Meldgaard, Kaggle forecasting competitions: An overlooked learning opportunity. Int. J. Forecast. 37, 587\u2013603 (2021).<\/span><\/section>\n
142. D. Colombo, M. H. Maathuis, Order-independent constraint-based causal structure<\/span><\/section>\n
learning. J. Mach. Learn. Res. 15, 3741\u20133782 (2014).<\/span><\/section>\n
143. P. Spirtes, C. Glymour, An algorithm for fast recovery of sparse causal graphs. Soc. Sci. Comput. Rev. 9, 62\u201372 (1991).<\/span><\/section>\n
144. D. Colombo, M. H. Maathuis, M. Kalisch, T. S. Richardson, Learning high-dimensional directed acyclic graphs with latent and selection variables. Ann. Stat. 40, 294\u2013321 (2012).<\/span><\/section>\n
145. I. Tsamardinos, C. F. Aliferis, A. Statnikov, Time and sample efficient discovery of Markov blankets and direct causal relations, in Proceedings of the Ninth ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (Association for Computing Machinery, 2003), pp. 673\u2013678.<\/span><\/section>\n
146. I. Tsamardinos, C. F. Aliferis, A. Statnikov, Algorithms for large scale Markov blanket<\/span><\/section>\n
discovery. FLAIRS Conf. 2, 376\u2013380 (2003b).<\/span><\/section>\n
147. M. Scutari, C. Vitolo, A. Tucker, Learning Bayesian networks from big data with greedy search: Computational complexity and efficient implementation. Stat. Comput. 29, 1095\u20131108 (2019).<\/span><\/section>\n
148. J. I. Alonso-Barba, J. A. G\u00e1mez, J. M. Puerta, Scaling up the greedy equivalence search algorithm by constraining the search space of equivalence classes. Int. J. Approx. Reason. 54, 429\u2013451 (2013).<\/span><\/section>\n
149. I. Tsamardinos, L. E. Brown, C. F. Aliferis, The max-min hill-climbing Bayesian network structure learning algorithm. Mach. Learn. 65, 31\u201378 (2006).<\/span><\/section>\n


<\/span><\/p>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n

\n
\n

\uff08\u53c2\u8003\u6587\u732e\u53ef\u4e0a\u4e0b\u6ed1\u52a8<\/strong>\u67e5\u770b\uff09<\/p>\n<\/section>\n<\/section>\n


<\/span><\/p>\n


<\/p>\n

\n
\n
\n
\n
\n

\u56e0\u679c\u8868\u5f81\u5b66\u4e60\u8bfb\u4e66\u4f1a\u542f\u52a8<\/strong><\/span><\/p>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n


<\/span><\/p>\n

<\/span><\/span><\/a><\/p>\n


<\/span><\/section>\n
\u8be6\u60c5\u8bf7\u89c1\uff1a<\/span><\/span><\/section>\n
\u8fde\u63a5\u56e0\u679c\u79d1\u5b66\u4e0e\u6df1\u5ea6\u5b66\u4e60\u7684\u6865\u6881\uff1a\u56e0\u679c\u8868\u5f81\u5b66\u4e60\u8bfb\u4e66\u4f1a\u542f\u52a8<\/a><\/span>
<\/span><\/section>\n

<\/section>\n

<\/section>\n
\n

\u63a8\u8350\u9605\u8bfb<\/span><\/strong><\/strong><\/p>\n