{"id":22045,"date":"2020-11-20T21:20:28","date_gmt":"2020-11-20T13:20:28","guid":{"rendered":"https:\/\/swarma.org\/?p=22045"},"modified":"2020-11-20T21:20:28","modified_gmt":"2020-11-20T13:20:28","slug":"%e3%80%8a%e5%9b%a0%e6%9e%9c%e7%a7%91%e5%ad%a6%e5%91%a8%e5%88%8a%e3%80%8b%e7%ac%ac2%e6%9c%9f%ef%bc%9a%e5%a6%82%e4%bd%95%e8%a7%a3%e5%86%b3%e6%b7%b7%e6%b7%86%e5%81%8f%e5%b7%ae%ef%bc%9f","status":"publish","type":"post","link":"https:\/\/swarma.org\/?p=22045","title":{"rendered":"\u300a\u56e0\u679c\u79d1\u5b66\u5468\u520a\u300b\u7b2c2\u671f\uff1a\u5982\u4f55\u89e3\u51b3\u6df7\u6dc6\u504f\u5dee\uff1f"},"content":{"rendered":"
\n
\n
\n
\n

\u4e3a\u4e86\u5e2e\u52a9\u5927\u5bb6\u66f4\u597d\u5730\u4e86\u89e3\u56e0\u679c\u79d1\u5b66\u7684\u6700\u65b0\u79d1\u7814\u8fdb\u5c55\u548c\u8d44\u8baf\uff0c\u6211\u4eec\u56e0\u679c\u79d1\u5b66\u793e\u533a\u56e2\u961f<\/span>\u672c\u5468\u6574\u7406\u4e86\u7b2c2\u671f\u300a\u56e0\u679c\u79d1\u5b66\u5468\u520a\u300b\uff0c\u4ece Causality, Causal Inference, Causal AI \u4e09\u4e2a\u7ef4\u5ea6\u9e1f\u77b0\uff0c\u63a8\u9001\u8fd1\u671f\u56e0<\/span>\u679c\u79d1\u5b66\u503c\u5f97\u5173\u6ce8\u7684\u8bba\u6587\u548c\u8d44\u8baf\u4fe1\u606f\uff0c \u540c\u65f6\u6211\u4eec\u4e5f\u5c06\u5411\u5927\u5bb6\u4ecb\u7ecd\u793e\u533a\u6b63\u5728\u63a8\u8fdb\u7684\u6d3b\u52a8\u2014\u2014\u56e0\u679c\u79d1\u5b66\u4e0eCausal AI\u8bfb\u4e66\u4f1a\u7b2c6\u671f\u4e2d\u7684\u4e3b\u8981\u62a5\u544a\u5185\u5bb9\u3001\u89c2\u70b9\u3002<\/span><\/p>\n

\u672c\u671f\u4f5c\u8005\uff1a\u51b5\u7428\uff0c\u9f9a\u9e64\u626c\uff0c\u9648\u6657\u66e6\uff0c\u9648\u5929\u8c6a\uff0c\u5f20\u5353\u5a67\uff0c\u6768\u96c5\u7a0b<\/em><\/p>\n<\/section>\n<\/section>\n<\/section>\n

\n

<\/section>\n<\/section>\n
\u672c\u671f\u5468\u62a5\u4e2d\u7684\u8bba\u6587\u63a8\u8350\uff0c\u5c06\u56f4\u7ed5\u56e0\u679c\u79d1\u5b66\u9886\u57df\u7684\u201c\u6df7\u6dc6\u504f\u5dee\u201d\u95ee\u9898\u5c55\u5f00\uff0c\u5173\u4e8e\u5b83\u7684\u89e3\u91ca\uff0c\u5927\u5bb6\u53ef\u4ee5\u5148\u770b\u4e0b\u9762\u8fd9\u4e2a\u4f8b\u5b50<\/span>\uff08\u719f\u6089\u7684\u670b\u53cb\u4e5f\u53ef\u4ee5\u5ffd\u7565\u8fd9\u90e8\u5206\u5185\u5bb9\uff0c\u76f4\u63a5\u9605\u8bfb\u4e0b\u9762\u7684\u201c\u8bba\u6587\u63a8\u8350\u201d\uff09<\/span>\u3002<\/span><\/section>\n

<\/section>\n
\u953b\u70bc\u80fd\u5426\u964d\u4f4e\u80c6\u56fa\u9187\u5462\uff1f\u5982\u4e0b\u56fe1\uff0c\u4ece\u6bcf\u4e2a\u5e74\u9f84\u5c42\u6765\u770b\u53ef\u4ee5\u964d\u4f4e\uff0c\u4f46\u662f\u5982\u679c\u4e0d\u5206\u5c42\u5219\u4f1a\u63d0\u9ad8\u80c6\u56fa\u9187\u3002<\/span><\/section>\n

<\/span><\/section>\n
<\/section>\n
\u56fe1: \u953b\u70bc\u662f\u5426\u6709\u5229\u4e8e\u5065\u5eb7\uff1f<\/span><\/section>\n

<\/section>\n
\u8fd9\u4e2a\u95ee\u9898\u4fbf\u6d89\u53ca\u6df7\u6dc6\u504f\u5dee\uff0c\u56de\u7b54\u5b83\u4ec5\u4ec5\u9760\u6570\u636e\u4e0d\u591f\uff0c\u9700\u8981\u56e0\u679c\u5efa\u6a21\uff0c\u8f6c\u5316\u672c\u671f\u5468\u520a\u5173\u6ce8\u7684\u56e0\u679c\u95ee\u9898\uff1a\u5728\u62e5\u6709\u6cbb\u7597\u53d8\u91cf T\uff0c\u534f\u53d8\u91cf X \u548c\u7ed3\u679c\u53d8\u91cf Y \u7684\u89c2\u6d4b\u6570\u636e\u4e0b\u7684\u56e0\u679c\u6548\u5e94\u4f30\u8ba1\u3002<\/span><\/section>\n

<\/span><\/section>\n
<\/section>\n
\u56fe2\uff1a\u56e0\u679c\u4e4b\u68af\u548c\u56e0\u679c\u63a8\u7406\u5f15\u64ce<\/span><\/section>\n

<\/section>\n
\u63a8\u7406\u5f15\u64ce\u4e2d\uff0c\u8be5\u95ee\u9898\u5c5e\u4e8e\u56e0\u679c\u4e4b\u68af\u5e72\u9884\u5c42\u7684 Query\uff0c\u800c Data \u662f\u89c2\u6d4b\u6570\u636e\uff0cAsumptions \u5219\u7ecf\u5e38\u7528\u6f5c\u7ed3\u679c\u6846\u67b6<\/span>\uff08Potential Outcome\uff09<\/span>\u6765\u63cf\u8ff0\u3002<\/span><\/section>\n

<\/section>\n
\u5173\u4e8e\u8be5\u56e0\u679c\u95ee\u9898\u5982\u4f55\u56de\u7b54\uff0c\u4e5f\u5c31\u662f\u53bb\u6df7\u6dc6\u504f\u5dee\uff0c\u6d59\u6c5f\u5927\u5b66\u52a9\u7406\u6559\u6388\u51b5\u7428\u5411\u5927\u5bb6\u63a8\u8350\u4e86\u51e0\u7bc7\u8bba\u6587\uff0c\u6211\u4eec\u6839\u636e\u201c\u57fa\u4e8e\u5339\u914d\u65b9\u6cd5\u201d\u201c\u503e\u5411\u8bc4\u5206\u65b9\u6cd5\u201d\u548c\u201c\u76f4\u63a5\u5747\u8861\u65b9\u6cd5\u201d\u4e09\u4e2a\u7c7b\u522b\u5206\u522b\u9009\u62e9\u4e24\u7bc7\u8bba\u6587\u8fdb\u884c\u4e86\u6574\u7406\u548c\u89e3\u8bfb\u3002<\/span><\/section>\n

<\/section>\n
1. \u8bba\u6587\u63a8\u8350<\/strong><\/span>
<\/span><\/section>\n

<\/section>\n
\u524d\u4e24\u7bc7\u8bba\u6587\u662f\u57fa\u4e8e\u5339\u914d\u7684\u65b9\u6cd5<\/span> (Matching based method) <\/span>\uff0c\u8be5\u65b9\u6cd5\u57fa\u672c\u601d\u60f3\u662f\u5bf9\u6bd4\u76f8\u4f3c\u4e2a\u4f53\u7528\u836f\u548c\u4e0d\u7528\u836f\u4ea7\u751f\u7684\u6548\u679c\u5dee\u5f02\u3002\u4e2d\u95f4\u4e24\u7bc7\u662f\u4ee5\u503e\u5411\u8bc4\u5206\u4e3a\u5de5\u5177\uff0c\u7ed9\u5b9a\u503e\u5411\u8bc4\u5206\u5219\u7c7b\u4f3c\u4e8e\u968f\u673a\u5316\u5b9e\u9a8c\uff0c\u800c\u6700\u540e\u4e24\u7bc7\u662f\u901a\u8fc7\u76f4\u63a5\u52a0\u6743\u521b\u9020\u4e00\u4e2a\u65b0\u7684\u603b\u4f53\uff0c\u4f7f\u5f97\u6df7\u6dc6\u53d8\u91cf\u548c\u6cbb\u7597\u53d8\u91cf\u72ec\u7acb\u7684\u65b9\u6cd5\u3002<\/span><\/section>\n

<\/section>\n
1.1 \u57fa\u4e8e\u5339\u914d\u7684\u65b9\u6cd5(Matching based method) \u4e00\u4e2a\u524d\u6cbf\u7406\u8bba\u6846\u67b6<\/span><\/strong><\/section>\n

<\/section>\n
Kallus, N. 2017. A Framework for Optimal Matching for Causal Inference. In Artificial Intelligence and Statistics, 372\u2013381.    <\/span><\/section>\n

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

\u8bba\u6587\u6807\u9898\uff1a\u56e0\u679c\u63a8\u65ad\u7684\u6700\u4f18\u5339\u914d\u6846\u67b6\u6846\u67b6<\/span><\/strong><\/p>\n

\u6458\u8981\uff1a<\/span><\/strong>\u672c\u6587\u63d0\u51fa\u4e86\u4e00\u79cd\u4ece\u89c2\u6d4b\u6570\u636e\u4e2d\u8fdb\u884c\u56e0\u679c\u63a8\u65ad\u7684\u5e7f\u4e49\u6700\u4f18\u5339\u914d\u65b9\u6cd5<\/span> (generalized optimal matching, GOM)<\/span>\uff0c\u5b83\u6db5\u76d6\u4e86 atching\u3001covariate balancing \u4ee5\u53ca doubly robust \u7b49\u65b9\u6cd5\u3002\u8fd9\u5957\u6846\u67b6\u662f\u57fa\u4e8e\u5bf9\u6700\u4f18\u5339\u914d\u7684\u4e00\u79cd\u65b0\u7684\u6cdb\u51fd\u5206\u6790\u7684\u63a8\u5e7f\u63d0\u51fa\u7684\uff0c\u5b83\u4ea7\u751f\u4e86\u4e00\u7c7b GOM \u7684\u65b9\u6cd5\uff0c\u672c\u6587\u63d0\u4f9b\u4e86\u4e00\u5957\u7edf\u4e00\u7684\u7406\u8bba\u6846\u67b6\u6765\u5bf9\u5b83\u4eec\u8fdb\u884c\u53ef\u89e3\u6027\u548c\u4e00\u81f4\u6027\u5206\u6790\u3002\u8bb8\u591a\u5df2\u6709\u7684\u65b9\u6cd5\u90fd\u53ef\u4ee5\u88ab\u7eb3\u5165 GOM \u7684\u6846\u67b6\uff0c\u5229\u7528GOM\u89c6\u89d2\u7684\u89e3\u91ca\uff0c\u53ef\u4ee5\u5c06\u5b83\u4eec\u62d3\u5c55\u6210\u4e00\u79cd\u6700\u4f18\u4e14\u81ea\u52a8\u7684\u65b9\u5dee\u4e0e\u6027\u80fd\u4e4b\u95f4\u7684\u5e73\u8861\u7b56\u7565\u3002Kernel optimal matching <\/span>(KOM) <\/span>\u4f5c\u4e3aGOM\u7684\u4e00\u7c7b\u5b50\u7c7b\uff0c\u7406\u8bba\u548c\u7ecf\u9a8c\u7ed3\u8bba\u8868\u660e\uff0c\u53ef\u4ee5\u5c06\u8bb8\u591a\u65b9\u6cd5\u7684\u4f18\u70b9\u6c47\u96c6\u5728\u8fd9\u4e00\u7c7b\u65b9\u6cd5\u4e2d\u3002KOM\u53ef\u4ee5\u8f6c\u5316\u4e3a\u6c42\u89e3\u7ebf\u6027\u7ea6\u675f\u7684\u51f8\u4e8c\u6b21\u4f18\u5316\u95ee\u9898\uff0c\u5728\u7ee7\u627f\u4e86\u53ef\u89e3\u91ca\u6027\u4e0e model-free \u7684\u5339\u914d\u4e00\u81f4\u6027\u540c\u65f6\uff0c\u8fd8\u5b9e\u73b0\u4e86\u5728\u7279\u5b9a\u56de\u5f52\u95ee\u9898\u4e0b\u7684<\/span>\u3001\u51cf\u5c11 bias \u4ee5\u53ca\u548c doubly robust \u65b9\u6cd5\u76f8\u5f53\u7684\u9c81\u68d2\u6027\u3002\u5728\u6709\u9650\u91cd\u53e0<\/span> (limited overlap) <\/span>\u7684\u8bbe\u5b9a\u4e0b\uff0cKOM\u662f\u4e00\u79cd\u5bf9\u4e8e\u90e8\u5206\u8bc6\u522b\u548c\u9c81\u68d2\u8986\u76d6\u95ee\u9898\u7684\u53ef\u79fb\u690d\u7684\u533a\u95f4\u4f30\u8ba1\u65b9\u6cd5\u3002\u6587\u7ae0\u5728\u751f\u6210\u6570\u636e\u548c\u771f\u5b9e\u6570\u636e\u4e0b\u9a8c\u8bc1\u4e86\u8fd9\u70b9\u3002<\/span><\/section>\n
<\/span><\/section>\n

<\/section>\n
Kallus, N. 2019. Generalized optimal matching methods for causal inference. The Journal of Machine Learning Research (forthcoming)   <\/span> <\/section>\n

<\/section>\n

\u8bba\u6587\u6807\u9898\uff1a\u56e0\u679c\u63a8\u65ad\u7684\u5e7f\u4e49\u6700\u4f18\u5339\u914d\u65b9\u6cd5<\/span><\/strong><\/span><\/p>\n

\u6458\u8981\uff1a<\/span><\/strong><\/span>\u672c\u6587\u552f\u4e00\u4f5c\u8005 <\/span>Nathan Kallus \u4e5f\u662f\u4e0a\u4e00\u7bc7\u63a8\u6587\u201cA framework for optimal matching for causal inference\u201d\u7684\u552f\u4e00\u4f5c\u8005\uff0c\u672c\u6587\u201cGeneralized Optimal Matching Methods for Causal Inference\u201d\u662f\u57fa\u4e8e\u4e0a\u4e00\u7bc7\u5de5\u4f5c\u63a8\u5e7f\u7684\u540e\u7eed\u5de5\u4f5c\uff0c\u6574\u4f53\u4e0a\u5ef6\u7eed\u4e86\u5148\u524d\u7684\u7814\u7a76\u601d\u8def\uff0c\u4f46\u662f\u7ed9\u51fa\u4e86\u66f4\u8be6\u5c3d\u7684\u7406\u8bba\u4f9d\u636e\u5e76\u63d0\u51fa\u4e86KOM++\u8fd9\u79cd\u65b0\u7684\u5339\u914d\u7b56\u7565\u3002<\/span>\u6587\u7ae0\u7684\u7406\u8bba\u6027\u540c\u6837\u5341\u5206\u5f3a\uff0c\u4f46\u4f5c\u8005\u4e5f\u5728KOM\u7ae0\u8282\u7ed9\u51fa\u4e86\u4e00\u4e9b\u8bf8\u5982 kernel \u9009\u62e9\u7b49\u5b9e\u8df5\u5316\u7684\u5efa\u8bae\u4e0e\u8ba8\u8bba\uff0c\u5341\u5206\u63a8\u8350\u5728\u56e0\u679c\u63a8\u65ad matching \u9886\u57df\u7684\u7814\u7a76\u8005\u9605\u8bfb\uff0c\u4e5f\u5efa\u8bae\u5bf9\u56e0\u679c\u63a8\u65ad\u3001\u673a\u5668\u5b66\u4e60\u7406\u8bba\u611f\u5174\u8da3\u7684\u670b\u53cb\u8fdb\u4e00\u6b65\u9605\u8bfb\u3002<\/span><\/section>\n

<\/section>\n
1.2 \u57fa\u4e8e\u503e\u5411\u8bc4\u5206\u7684\u65b9\u6cd5 (Propensity score based method)\uff0c\u4e00\u7bc7\u7efc\u8ff0\u548c\u4e00\u7bc7\u524d\u6cbf<\/span><\/strong><\/section>\n

<\/span><\/strong><\/section>\n
Austin, P. C. 2011. An introduction to propensity score methods for reducing the effects of confounding in observational studies. Multivariate behavioral research 46(3): 399\u2013424.   <\/span><\/section>\n

<\/section>\n

\u8bba\u6587\u6807\u9898\uff1a\u5728\u89c2\u6d4b\u7814\u7a76\u4e2d\u7528\u4e8e\u51cf\u5c11\u6df7\u6dc6\u53d8\u91cf\u5f71\u54cd\u7684\u503e\u5411\u6027\u8bc4\u5206\u65b9\u6cd5\u7b80\u4ecb<\/span><\/strong><\/span><\/p>\n

\u6458\u8981\uff1a<\/span><\/strong><\/span>\u503e\u5411\u6027\u8bc4\u5206\u662f\u7ed9\u5b9a\u89c2\u6d4b\u7279\u5f81\u6761\u4ef6\u4e0b\u7684\u63a5\u53d7\u6cbb\u7597\u6982\u7387\u8d4b\u503c\uff0c\u5b83\u901a\u8fc7\u6a21\u4eff\u968f\u673a\u5316\u5b9e\u9a8c\u7684\u4e00\u4e9b\u7279\u5b9a\u7279\u5f81\u6765\u5141\u8bb8\u7814\u7a76\u8005\u8fdb\u884c\u89c2\u6d4b\u7814\u7a76\u7684\u8bbe\u8ba1\u548c\u5206\u6790\u3002<\/span>\u5177\u4f53\u800c\u8a00\uff0c\u503e\u5411\u6027\u8bc4\u5206\u662f\u4e00\u79cd\u5e73\u8861\u8bc4\u5206\uff1a<\/span>\u5728\u7ed9\u5b9a\u503e\u5411\u6027\u8bc4\u5206\u60c5\u51b5\u4e0b\uff0c\u89c2\u6d4b\u5230\u7684\u534f\u53d8\u91cf\u5206\u5e03\u4f1a\u8fd1\u4f3c\u4e8e\u968f\u673a\u5316\u5b9e\u9a8c\u7684\u5206\u5e03\u3002<\/span>\u672c\u6587\u8ba8\u8bba\u4e86\u56db\u79cd\u503e\u5411\u6027\u8bc4\u5206\u65b9\u6cd5\uff1a<\/span>\u57fa\u4e8e\u503e\u5411\u6027\u8bc4\u5206\u7684\u5339\u914d\u6cd5\u3001\u57fa\u4e8e\u503e\u5411\u6027\u8bc4\u5206\u7684\u5206\u5c42\u6cd5\u3001\u57fa\u4e8e\u503e\u5411\u6027\u8bc4\u5206\u7684 inverse probability of treatment <\/span><\/span>weighting(IPW) \u6cd5\u4ee5\u53ca\u57fa\u4e8e\u503e\u5411\u6027\u8bc4\u5206\u7684\u534f\u53d8\u91cf\u8c03\u6574\u6cd5\u3002\u672c\u6587\u63cf\u8ff0\u4e86\u4e00\u79cd\u5e73\u8861\u8bca\u65ad\u7a0b\u5e8f\uff0c\u7528\u4e8e\u68c0\u9a8c\u4f7f\u7528\u7684\u503e\u5411\u6027\u8bc4\u5206\u65b9\u6cd5\u662f\u5426\u5408\u7406\u3002\u6b64\u5916\uff0c\u672c\u6587\u8fd8\u8ba8\u8bba\u4e86\u57fa\u4e8e\u56de\u5f52\u7684\u65b9\u6cd5\u548c\u57fa\u4e8e\u503e\u5411\u6027\u8bc4\u5206\u7684\u65b9\u6cd5\u5728\u89c2\u6d4b\u6570\u636e\u5206\u6790\u4e0a\u7684\u533a\u522b\u3002\u6587\u7ae0\u6700\u540e\u63cf\u8ff0\u4e86\u4e0d\u540c\u7684\u5e73\u5747\u56e0\u679c\u6548\u5e94\u4e0e\u503e\u5411\u6027\u8bc4\u5206\u5206\u6790\u7684\u8054\u7cfb\u3002<\/span><\/section>\n

<\/section>\n
Kun Kuang, Peng Cui, Hao Zou, Bo Li, Jianrong Tao, Fei Wu, and Shiqiang Yang. Data-Driven Variable Decomposition for Treatment Effect Estimation, IEEE Transaction on Knowledge and Data Engineering (TKDE) , 2020    <\/span><\/section>\n

<\/section>\n

\u8bba\u6587\u6807\u9898\uff1a\u6570\u636e\u9a71\u52a8\u7684\u53d8\u91cf\u5206\u89e3\u7528\u4e8e\u56e0\u679c\u6548\u5e94\u4f30\u8ba1<\/span><\/strong><\/span><\/p>\n

\u6458\u8981\uff1a<\/span><\/strong>\u56e0\u679c\u63a8\u65ad\u7684\u4e00\u4e2a\u57fa\u672c\u95ee\u9898\u662f\u89c2\u5bdf\u7814\u7a76\u4e2d\u5b58\u5728\u6df7\u6dc6\u53d8\u91cf\u65f6\u7684\u56e0\u679c\u6548\u5e94\u4f30\u8ba1\u3002\u503e\u5411\u6027\u8bc4\u5206\u5e38\u88ab\u7528\u4e8e\u6df7\u6dc6\u6548\u5e94\u7684\u63a7\u5236\u3002\u4f46\u5b83\u5c06\u6240\u6709\u89c2\u5bdf\u5230\u7684\u53d8\u91cf\u89c6\u4e3a\u6df7\u6dc6\u53d8\u91cf\uff0c\u4ece\u800c\u5ffd\u7565\u4e86\u90a3\u4e9b\u5bf9\u5904\u7406\u6ca1\u6709\u5f71\u54cd\uff0c\u4f46\u5bf9\u4e8e\u7ed3\u679c\u5177\u6709\u9884\u6d4b\u6027\u7684\u8c03\u6574\u53d8\u91cf\u3002\u6700\u8fd1\u7814\u7a76\u8bc1\u660e\uff0c\u8c03\u6574\u53d8\u91cf\u53ef\u4ee5\u6709\u6548\u51cf\u5c11\u4f30\u8ba1\u56e0\u679c\u6548\u5e94\u7684\u65b9\u5dee\u3002\u7136\u800c\uff0c\u5982\u4f55\u81ea\u52a8\u5206\u79bb\u6df7\u6dc6\u53d8\u91cf\u548c\u8c03\u6574\u53d8\u91cf\u4f9d\u7136\u662f\u4e00\u4e2a\u5f00\u653e\u6027\u95ee\u9898\u3002\u5728\u8fd9\u7bc7\u6587\u7ae0\u4e2d\uff0c\u6211\u4eec\u9996\u6b21\u63d0\u51fa\u4e00\u79cd\u6570\u636e\u9a71\u52a8\u7684\u53d8\u91cf\u5206\u89e3 <\/span>(Data-Driven Variable Decomposition, D2VD) <\/span>\u7b97\u6cd5\uff0c\u5b83\u53ef\u4ee5\u81ea\u52a8\u5c06\u53d8\u91cf\u5206\u79bb\u4e3a\u6df7\u6dc6\u53d8\u91cf\u548c\u8c03\u6574\u53d8\u91cf\uff0c\u5e76\u540c\u6b65\u5730\u4f30\u8ba1\u56e0\u679c\u6548\u5e94\u3002\u5728\u6807\u51c6\u5047\u8bbe\u4e0b\uff0c\u6211\u4eec\u4ece\u7406\u8bba\u4e0a\u8bc1\u660e\u4e86D2VD \u7b97\u6cd5\u80fd\u4ee5\u66f4\u4f4e\u7684\u65b9\u5dee\u7ed9\u51fa\u56e0\u679c\u6548\u5e94\u7684\u65e0\u504f\u4f30\u8ba1\u3002\u6b64\u5916\uff0c\u4e3a\u4e86\u89e3\u51b3\u975e\u7ebf\u6027\u95ee\u9898\uff0c\u6211\u4eec\u63d0\u51fa\u4e86\u4e00\u79cd\u975e\u7ebf\u6027\u7684D2VD<\/span> (Nonlinear-D2VD, N-D2VD) <\/span>\u7b97\u6cd5\u3002\u4e3a\u4e86\u9a8c\u8bc1\u7b97\u6cd5\u7684\u6709\u6548\u6027\uff0c\u6211\u4eec\u5728\u5408\u6210\u6570\u636e\u96c6\u548c\u771f\u5b9e\u6570\u636e\u96c6\u4e0a\u8fdb\u884c\u4e86\u5927\u91cf\u7684\u5b9e\u9a8c\u3002\u5b9e\u9a8c\u7ed3\u679c\u8868\u660e\uff0c\u4e0e\u73b0\u6709\u7684\u65b9\u6cd5\u76f8\u6bd4\uff0cD2VD \u548c N-D2VD \u7b97\u6cd5\u80fd\u591f\u81ea\u52a8\u800c\u7cbe\u786e\u5730\u5206\u79bb\u53d8\u91cf\uff0c\u66f4\u51c6\u786e\u5730\u4f30\u8ba1\u56e0\u679c\u6548\u5e94\uff0c\u4e14\u65b9\u5dee\u66f4\u5c0f\u3002\u6211\u4eec\u8fd8\u8868\u660e\uff0c\u5728\u4e00\u4e2a\u5728\u7ebf\u5e7f\u544a\u6570\u636e\u96c6\u4e2d\uff0c\u6211\u4eec\u7684\u7b97\u6cd5\u4ea7\u751f\u6392\u540d\u9760\u524d\u7684\u7279\u5f81\u5177\u6709\u6700\u597d\u7684\u9884\u6d4b\u6027\u80fd\u3002<\/span>
<\/section>\n

<\/section>\n
1.3 \u76f4\u63a5\u6df7\u6dc6\u56e0\u5b50\u5747\u8861\u65b9\u6cd5 (Directly confounder balancing)<\/span><\/strong><\/span><\/section>\n

<\/section>\n
K. Imai and M. Ratkovic. Covariate balancing propensity score. Journal of the Royal <\/span>Statistical Society: Series B (Statistical Methodology), 76(1):243\u2013263, 2014.    <\/span>
<\/section>\n

<\/section>\n

\u8bba\u6587\u6807\u9898\uff1a\u534f\u53d8\u91cf\u5747\u8861\u7684\u503e\u5411\u8bc4\u5206<\/span><\/strong><\/span><\/p>\n

\u6458\u8981\uff1a<\/span><\/strong>\u503e\u5411\u8bc4\u5206\u5728\u5404\u79cd\u56e0\u679c\u63a8\u65ad\u4e2d\u626e\u6f14\u7740\u6838\u5fc3\u89d2\u8272\u3002\u7279\u522b\u5730\uff0c\u5728\u89c2\u5bdf\u6027\u6570\u636e\u7684\u5206\u6790\u4e2d\uff0c\u57fa\u4e8e\u503e\u5411\u8bc4\u5206\u4f30\u8ba1\u7684\u5339\u914d\u548c\u52a0\u6743\u65b9\u6cd5\u6108\u53d1\u5e38\u89c1\u3002\u5c3d\u7ba1\u8fd9\u4e9b\u65b9\u6cd5\u5e7f\u53d7\u6b22\u8fce\uff0c\u800c\u4e14\u5728\u7406\u8bba\u4e0a\u5177\u6709\u5438\u5f15\u529b\uff0c\u4f46\u662f\u5b83\u4eec\u5b9e\u9645\u56f0\u96be\u4e3b\u8981\u662f\u5fc5\u987b\u4f30\u8ba1\u503e\u5411\u8bc4\u5206\u3002\u7814\u7a76\u8005\u5df2\u7ecf\u53d1\u73b0\uff0c\u5bf9\u503e\u5411\u8bc4\u5206\u6a21\u578b\u7684\u5fae\u5c0f\u8bef\u5224\u4f1a\u5bfc\u81f4\u56e0\u679c\u6548\u5e94\u4f30\u8ba1\u7684\u4e25\u91cd\u504f\u5dee\u3002\u6211\u4eec\u5f15\u5165\u534f\u53d8\u91cf\u5747\u8861\u7684\u503e\u5411\u8bc4\u5206<\/span> (Covariate Balancing Propensity Score, CBPS) <\/span>\u65b9\u6cd5\uff0c\u5728\u5bf9 Treatment \u8d4b\u503c\u8fdb\u884c\u5efa\u6a21\u7684\u540c\u65f6\uff0c\u4f18\u5316\u534f\u53d8\u91cf\u5747\u8861\u6027\u3002\u4e5f\u5c31\u662f\u8bf4 CBPS \u540c\u65f6\u5229\u7528\u503e\u5411\u8bc4\u5206\u5e2e\u52a9\u534f\u53d8\u91cf\u5747\u8861\u548c\u5efa\u6a21 Treatment \u8d4b\u503c\u6761\u4ef6\u6982\u7387\u3002CBPS \u7684\u4f30\u8ba1\u53ef\u4ee5\u7528\u5e7f\u4e49\u77e9\u4f30\u8ba1\u6216\u8005\u7ecf\u9a8c\u4f3c\u7136\u6846\u67b6\u5b9e\u73b0\u3002\u6211\u4eec\u53d1\u73b0 CBPS \u663e\u8457\u6539\u5584\u4e86\u503e\u5411\u8bc4\u5206\u5339\u914d\u548c\u52a0\u6743\u65b9\u6cd5\u5728\u6587\u732e\u62a5\u9053\u4e2d\u7cdf\u7cd5\u7684\u5b9e\u8bc1\u8868\u73b0\u3002\u6211\u4eec\u8fd8\u8868\u660e\uff0cCBPS \u53ef\u4ee5\u63a8\u5e7f\u5230\u5176\u4ed6\u91cd\u8981\u7684\u73af\u5883\u4e2d\uff0c\u5305\u62ec\u4f30\u8ba1\u975e\u4e8c\u503c\u5904\u7406\u7684\u5e7f\u4e49\u503e\u5411\u8bc4\u5206\u4ee5\u53ca\u5c06\u5b9e\u9a8c\u4f30\u8ba1\u503c\u63a8\u5e7f\u5230\u76ee\u6807\u4eba\u7fa4\u3002\u6211\u4eec\u63d0\u4f9b\u4e86\u4e00\u4e2a\u5f00\u6e90\u8f6f\u4ef6\u7528\u4e8e\u5b9e\u73b0\u4e0a\u8ff0\u63d0\u51fa\u7684\u65b9\u6cd5\u3002<\/span>
<\/section>\n
<\/span><\/section>\n

<\/section>\n
Kun Kuang, Peng Cui, Bo Li, Meng Jiang, Fei Wu and Shiqiang Yang. Treatment Effect Estimation via Differentiated Confounder Balancing and Regression, Transactions on Knowledge Discovery from Data (TKDD) , 2019.<\/span><\/section>\n

<\/section>\n

\u8bba\u6587\u6807\u9898\uff1a\u901a\u8fc7\u533a\u5206\u6027\u6df7\u6dc6\u53d8\u91cf\u5747\u8861\u548c\u56de\u5f52\u5f97\u5230\u56e0\u679c\u6548\u5e94\u4f30\u8ba1<\/span><\/strong><\/span><\/p>\n

\u6458\u8981\uff1a<\/span><\/strong>\u56e0\u679c\u6548\u5e94\u5728\u8bf8\u5982\u793e\u4f1a\u8425\u9500\u3001\u533b\u7597\u4fdd\u5065\u548c\u516c\u5171\u653f\u7b56\u7b49\u9886\u57df\u7684\u51b3\u7b56\u4e2d\u626e\u6f14\u7740\u91cd\u8981\u89d2\u8272\u3002\u5728\u4e00\u822c\u7684\u89c2\u5bdf\u6027\u7814\u7a76\u4e2d\uff0c\u4f30\u8ba1\u56e0\u679c\u6548\u5e94\u7684\u5173\u952e\u6311\u6218\u662f\u63a7\u5236\u7531\u5904\u7406\u5355\u5143\u548c\u5bf9\u7167\u5355\u5143\u4e4b\u95f4\u6df7\u6dc6\u53d8\u91cf\u5206\u5e03\u4e0d\u5747\u8861\u5f15\u8d77\u7684\u6df7\u6dc6\u504f\u5dee\u3002\u4f20\u7edf\u7684\u65b9\u6cd5\u5728\u65e0\u6df7\u6dc6\u6027\u5047\u8bbe\u4e0b\uff0c\u7528\u5047\u5b9a\u662f\u51c6\u786e\u7684\u503e\u5411\u8bc4\u5206\u4f30\u8ba1\u6765\u91cd\u65b0\u52a0\u6743\u5355\u5143\uff0c\u4ee5\u6d88\u9664\u6df7\u6dc6\u504f\u5dee\u3002\u63a7\u5236\u9ad8\u7ef4\u53d8\u91cf\u53ef\u4ee5\u4f7f\u65e0\u6df7\u6dc6\u6027\u5047\u8bbe\u66f4\u52a0\u53ef\u4fe1\uff0c\u4f46\u5374\u5728\u51c6\u786e\u4f30\u8ba1\u503e\u5411\u8bc4\u5206\u4e0a\u4ea7\u751f\u4e86\u65b0\u6311\u6218\u3002\u6700\u8fd1\u7684\u4e00\u7cfb\u5217\u6587\u732e\u5e0c\u671b\u8df3\u8fc7\u503e\u5411\u8bc4\u5206\u4f30\u8ba1\uff0c\u76f4\u63a5\u4f18\u5316\u6743\u91cd\u6765\u5747\u8861\u6df7\u6dc6\u53d8\u91cf\u7684\u5206\u5e03\u3002\u4f46\u662f\u5f53\u524d\u7684\u5747\u8861\u65b9\u6cd5\u65e0\u6cd5\u5728\u5927\u91cf\u6f5c\u5728\u7684\u6df7\u6dc6\u53d8\u91cf\u505a\u51fa\u9009\u62e9\u548c\u533a\u5206\uff0c\u5bfc\u81f4\u5728\u8bb8\u591a\u9ad8\u7ef4\u73af\u5883\u4e2d\u53ef\u80fd\u8868\u73b0\u4e0d\u4f73\u3002\u5728\u8fd9\u7bc7\u6587\u7ae0\u4e2d\uff0c\u6211\u4eec\u63d0\u51fa\u4e86\u4e00\u4e2a\u6570\u636e\u9a71\u52a8\u7684\u533a\u5206\u6027\u6df7\u6dc6\u53d8\u91cf\u5747\u8861<\/span> (Differentiated Confounder Balancing, DCB) <\/span>\u7b97\u6cd5\uff0c\u6765\u8054\u5408\u9009\u62e9\u6df7\u6dc6\u53d8\u91cf\u3001\u533a\u5206\u6df7\u6dc6\u53d8\u91cf\u6743\u91cd\u548c\u5747\u8861\u6df7\u6dc6\u53d8\u91cf\u7684\u5206\u5e03\uff0c\u4ee5\u5728\u9ad8\u7ef4\u73af\u5883\u4e0b\u5b9e\u73b0\u56e0\u679c\u6548\u5e94\u4f30\u8ba1\u3002\u6b64\u5916\uff0c\u5728\u4e00\u4e9b\u5b58\u5728\u4e25\u91cd\u6df7\u6dc6\u504f\u5dee\u7684\u60c5\u51b5\u4e0b\uff0c\u4e3a\u4e86\u8fdb\u4e00\u6b65\u51cf\u5c0f\u56e0\u679c\u6548\u5e94\u4f30\u8ba1\u7684\u504f\u5dee\u548c\u65b9\u5dee\uff0c\u6211\u4eec\u63d0\u51fa\u4e00\u79cd\u56de\u5f52\u6821\u6b63\u7684\u533a\u5206\u6027\u6df7\u6dc6\u53d8\u91cf\u5747\u8861<\/span> (Regression Adjusted Differentiated Confounder Balancing, RA-DCB) <\/span>\u7b97\u6cd5\uff0c\u8fd9\u79cd\u7b97\u6cd5\u57fa\u4e8e\u6211\u4eec\u7684DCB\u7b97\u6cd5\uff0c\u5e76\u7eb3\u5165\u4e86\u7ed3\u679c\u56de\u5f52\u6821\u6b63\u3002\u6211\u4eec\u63d0\u51fa\u7684\u534f\u540c\u5b66\u4e60\u7b97\u6cd5\u66f4\u80fd\u51cf\u5c11\u8bb8\u591a\u89c2\u5bdf\u6027\u7814\u7a76\u4e2d\u7684\u6df7\u6dc6\u504f\u5dee\u3002\u4e3a\u4e86\u9a8c\u8bc1\u4e0a\u8ff0DCB\u7b97\u6cd5\u548cRA-DCB\u7b97\u6cd5\u7684\u6709\u6548\u6027\uff0c\u6211\u4eec\u5728\u5408\u6210\u6570\u636e\u96c6\u548c\u771f\u5b9e\u4e16\u754c\u6570\u636e\u96c6\u4e2d\u8fdb\u884c\u4e86\u5927\u91cf\u5b9e\u9a8c\u3002\u5b9e\u9a8c\u7ed3\u679c\u6e05\u695a\u8868\u660e\u6211\u4eec\u7684\u7b97\u6cd5\u6bd4\u5f53\u4e0b\u6d41\u884c\u7684\u65b9\u6cd5\u5177\u6709\u66f4\u597d\u7684\u8868\u73b0\u3002\u901a\u8fc7\u7eb3\u5165\u56de\u5f52\u6821\u6b63\uff0c\u6211\u4eec\u7684RA-DCB\u7b97\u6cd5\u4f30\u8ba1\u5f97\u5230\u7684\u56e0\u679c\u6548\u5e94\u6bd4DCB\u7b97\u6cd5\u5f97\u5230\u7684\u5177\u6709\u66f4\u9ad8\u7684\u7cbe\u786e\u5ea6\uff0c\u7279\u522b\u662f\u5728\u4e25\u91cd\u6df7\u6dc6\u504f\u5dee\u7684\u60c5\u51b5\u4e0b\u3002\u6700\u540e\uff0c\u6211\u4eec\u5c55\u793a\u4e86\u7531\u6211\u4eec\u7b97\u6cd5\u6240\u4ea7\u751f\u7684\u6392\u540d\u9760\u524d\u7684\u7279\u5f81\u53ef\u4ee5\u51c6\u786e\u5730\u9884\u6d4b\u5728\u7ebf\u5e7f\u544a\u7684\u6548\u679c\u3002<\/span>
<\/section>\n
<\/span><\/section>\n

<\/section>\n
\u672c\u6b21\u63a8\u8350\u7684\u8bba\u6587\u4e3b\u8981\u5c5e\u4e8e Causal Inference for Data Science \uff0c\u4e03\u4e2a\u56e0\u679c\u63a8\u65ad\u5de5\u5177\u4e2d\u7684\u7b2c\u4e8c\u4e2a The Control of Confounding \u6709\u5173\u5185\u5bb9\u3002   <\/span><\/section>\n

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

<\/section>\n
\u56fe3\uff1a\u4e03\u4e2a\u56e0\u679c\u63a8\u65ad\u5de5\u5177<\/span><\/section>\n

<\/section>\n
2. \u8fd1\u671f\u8d44\u8baf\u63a8\u8350<\/span><\/strong><\/span><\/section>\n

<\/section>\n
\u6211\u4eec\u672c\u6b21\u7684\u8d44\u8baf\u63a8\u8350\u5305\u62ec\u4e00\u95e8\u65b0\u51fa\u7684\u56e0\u679c\u8bfe\u7a0b\uff0c\u4e24\u4e2a Causal AI \u65b9\u9762\u62a5\u544a\uff0c\u4e00\u7bc7\u56e0\u679c\u8fc1\u79fb\u5b66\u4e60\u7684\u8bba\u6587<\/span>\uff08\u6b63\u662f\u6211\u4eec\u8bfb\u4e66\u4f1a\u672c\u5468\u5468\u672b\u7684\u5c06\u8981\u5206\u4eab\u4e3b\u9898\uff09<\/span>\u3002<\/span><\/section>\n

<\/section>\n

2.1 Causal AI\u8bfe\u7a0b<\/span><\/strong><\/p>\n

\u8bfe\u7a0b\u540d\uff1aIntroduction to Causal Inference (ICI) <\/span>from a Machine Learning Perspective<\/span><\/strong>
<\/section>\n
<\/span><\/strong><\/section>\n

<\/span><\/section>\n
\u8fd9\u95e8\u8bfe\u7a0b\u7531 Yoshua Bengio \u9ad8\u5f92 Brady Neal <\/span>\u4e3b\u8bb2\uff0c\u4e3b\u8981\u8bb2\u8ff0\u56e0\u679c\u63a8\u7406\u76f8\u5173\u77e5\u8bc6\u3002<\/span>\u6b64\u5916\uff0c\u8be5\u8bfe\u7a0b\u6574\u5408\u4e86\u6765\u81ea\u8bb8\u591a\u4e0d\u540c\u9886\u57df\u7684\u89c1\u89e3\uff0c\u5982\u6d41\u884c\u75c5\u5b66\u3001\u7ecf\u6d4e\u5b66\u3001\u653f\u6cbb\u5b66\u548c\u673a\u5668\u5b66\u4e60\u7b49\uff0c\u8fd9\u4e9b\u9886\u57df\u90fd\u5229\u7528\u5230\u4e86\u56e0\u679c\u63a8\u7406\u3002<\/span>
<\/section>\n

<\/section>\n
\u8bfe\u7a0b\u94fe\u63a5\uff1ahttps:\/\/www.bradyneal.com\/causal-inference-course#course-textbook<\/span><\/section>\n

<\/span><\/section>\n
\u3010\u5185\u5bb9\u6807\u7b7e\u3011Causality \u56e0\u679c\u63a8\u7406\u57fa\u7840<\/span><\/strong><\/em><\/span><\/section>\n

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

2.2 Causal AI\u62a5\u544a<\/span><\/strong><\/p>\n

\u62a5\u544a\u540d\uff1aSymbolic, Statistical and Causal Artificial Intelligence<\/span><\/strong><\/span><\/section>\n
\u5728 MLSS2020 \u4e0a\uff0cBernhard Scholkopf \u9996\u5148\u7b80\u5355\u4ecb\u7ecd\u4e86\u8be5\u673a\u5668\u5b66\u4e60\u6691\u671f\u5b66\u6821\uff0c\u5b83\u5c06\u4f1a\u6d89\u53ca\u4ece\u57fa\u7840\u5230 state-of-art \u7684\u73b0\u4ee3\u673a\u5668\u5b66\u4e60\u6838\u5fc3\u4e3b\u9898\u3002<\/span>\u7136\u540e\u56de\u987e\u4e86\u4eba\u5de5\u667a\u80fd\u7684\u5386\u53f2\uff0c\u6307\u51fa Causality \u5c06\u4f1a\u662f\u4e0b\u4e00\u4ee3\u4eba\u5de5\u667a\u80fd\u7684\u5173\u952e\u3002<\/span>
<\/section>\n

<\/span><\/section>\n
<\/section>\n
Bernhard Scholkopf \u8bb2\u5ea7 Causal AI<\/span><\/section>\n

<\/section>\n
\u8bfe\u7a0b\u94fe\u63a5\uff1ahttps:\/\/www.youtube.com\/watch?v=8staJlMbAig<\/span><\/section>\n

<\/section>\n
\u3010\u5185\u5bb9\u6807\u7b7e\u3011Causal AI \u6982\u89c8<\/span><\/strong><\/em><\/section>\n

<\/section>\n

\u62a5\u544a\u540d\uff1a\u56e0\u679c\u5f3a\u5316\u5b66\u4e60\uff08Causal Reinforcement Learning\uff09<\/span><\/strong><\/p>\n

\u5728ICML2020\u4e0aElias Bareinboim\u6559\u6388\u7ec4\u7ec7\u4e86\u5173\u4e8e\u56e0\u679c\u5f3a\u5316\u5b66\u4e60\u7684 Tutorial, \u4ecb\u7ecd\u4e86\u56e0\u679c\u548c\u5f3a\u5316\u5b66\u4e60\u4e4b\u95f4\u7684\u8054\u7cfb\uff0c\u5e76\u4e14\u603b\u7ed3\u4e86\u56e0\u679c\u5f3a\u5316\u5b66\u4e60\u4e2d\u76846\u4e2a\u91cd\u8981\u7684\u4efb\u52a1\uff1a\u5c06\u5728\u7ebf\u5b66\u4e60\u548c\u79bb\u7ebf\u5b66\u4e60\u7ed3\u5408\uff0c\u5728\u5f3a\u5316\u5b66\u4e60\u4e2d\u52a0\u5165\u5408\u9002\u7684\u5e72\u9884\uff0c\u53cd\u4e8b\u5b9e\u51b3\u7b56\uff0c\u4f7f\u7528\u5f3a\u5316\u5b66\u4e60\u63d0\u53d6\u56e0\u679c\u6a21\u578b\uff0c\u4ee5\u53ca\u5728 Reward \u672a\u77e5\u7684\u60c5\u5f62\u4e0b\u8bad\u7ec3\u5f3a\u5316\u5b66\u4e60\u6a21\u578b\u3002<\/span><\/section>\n

<\/section>\n
<\/section>\n
\u89c1 ICML2020 https:\/\/crl.causalai.net\/<\/span><\/section>\n

<\/section>\n
\u3010\u5185\u5bb9\u6807\u7b7e\u3011Causal AI \u56e0\u679c\u673a\u5668\u5b66\u4e60<\/span><\/em><\/strong><\/section>\n

<\/section>\n
2.3 \u56e0\u679c\u8fc1\u79fb\u5b66\u4e60\u8bba\u6587<\/span><\/strong><\/section>\n

<\/section>\n
E<\/span>dmonds M, Ma X, Qi S, et al. Theory-Based Causal Transfer: Integrating Instance-Level Induction and Abstract-Level Structure Learning[C]\/\/AAAI. 2020: 1283-1291.<\/span><\/span><\/section>\n

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

\u8bba\u6587\u6807\u9898\uff1a\u57fa\u4e8e\u7406\u8bba\u7684\u56e0\u679c\u8fc1\u79fb\uff1a\u5b9e\u4f8b\u7ea7\u522b\u7684\u5f52\u7eb3\u53ca\u62bd\u8c61\u7ea7\u522b\u7684\u7ed3\u6784\u5b66\u4e60<\/span><\/strong><\/span><\/p>\n

\u6458\u8981\uff1a<\/span><\/strong><\/span>\u5728\u76f8\u8fd1\u4f46\u4e0d\u540c\u7684\u8bbe\u5b9a\u95f4\u5b66\u4e60\u53ef\u8fc1\u79fb\u7684\u77e5\u8bc6\u662f\u901a\u7528\u667a\u80fd\u7684\u57fa\u672c\u7ec4\u6210\u3002<\/span>\u672c\u6587\u4ece\u56e0\u679c\u7406\u8bba\u7684\u89c6\u89d2\u6765\u903c\u8fd1\u8fc1\u79fb\u5b66\u4e60\u7684\u6311\u6218\u3002<\/span>\u672c\u6587\u7684\u667a\u80fd\u4f53\u88ab\u8d4b\u4e88\u4e24\u6761\u57fa\u7840\u4f46\u4e00\u822c\u6027\u7684\u7406\u8bba\u6765\u8fdb\u884c\u8fc1\u79fb\u5b66\u4e60\uff1a<\/span>(i)\u8de8\u57df\u7684\u4efb\u52a1\u95f4\u6709\u4e00\u4e2a\u4e0d\u53d8\u7684\u4e00\u822c\u6027\u62bd\u8c61\u7ed3\u6784\uff1b<\/span>(ii)\u73af\u5883\u8868\u73b0\u51fa\u7684\u7279\u5b9a\u7279\u5f81\u5728\u8de8\u57df\u65f6\u7ef4\u6301\u5e38\u6570\u3002<\/span>\u672c\u6587\u91c7\u7528\u4e86\u8d1d\u53f6\u65af\u89c6\u89d2\u7684\u56e0\u679c\u7406\u8bba\u8fdb\u884c\u5f52\u7eb3\uff0c\u5e76\u7528\u8fd9\u4e9b\u7406\u8bba\u5728\u4e0d\u540c\u73af\u5883\u95f4\u6765\u8fc1\u79fb\u77e5\u8bc6\u3002<\/span>\u7ed9\u5b9a\u8fd9\u4e9b\u4e00\u822c\u6027\u7406\u8bba\uff0c\u672c\u6587\u7684\u76ee\u6807\u662f\u8bad\u7ec3\u4e00\u4e2a\u53ef\u548c\u95ee\u9898\u7a7a\u95f4\u4ea4\u4e92\u5e76\u63a2\u7d22\u7684\u667a\u80fd\u4f53\u6765\uff1a<\/span>(i)\u53d1\u73b0\u3001\u6784\u5efa\u5e76\u8fc1\u79fb\u6709\u7528\u7684\u62bd\u8c61\u7ed3\u6784\u5316\u77e5\u8bc6\uff1b<\/span>(ii)\u4ece\u73af\u5883\u4e2d\u89c2\u6d4b\u5230\u7684\u5b9e\u4f8b\u7ea7\u522b\u5c5e\u6027\u4e2d\u5f52\u7eb3\u51fa\u6709\u7528\u7684\u77e5\u8bc6\u3002<\/span>\u8d1d\u53f6\u65af\u7ed3\u6784\u7684\u5c42\u7ea7\u88ab\u7528\u4e8e\u5efa\u6a21\u62bd\u8c61\u5c42\u9762\u7684\u7ed3\u6784\u5316\u56e0\u679c\u77e5\u8bc6\uff0c\u5b9e\u4f8b\u7ea7\u522b\u7684\u76f8\u5173\u6027\u5b66\u4e60\u673a\u5236\u901a\u8fc7\u4ea4\u4e92\u6765\u5b66\u4e60\u54ea\u79cd\u7279\u5b9a\u76ee\u6807\u53ef\u4ee5\u88ab\u7528\u4e8e\u5f52\u7eb3\u72b6\u6001\u7684\u6539\u53d8\u3002<\/span>\u8fd9\u79cd\u6a21\u578b\u5b66\u4e60\u673a\u5236\u548c\u4e00\u4e2a\u57fa\u4e8e\u6a21\u578b\u7684\u89c4\u5212\u5668\u7ed3\u5408\u6765\u5b8c\u6210\u201c\u5f00\u9501\u201d\u73af\u5883\u4e2d\u7684\u4efb\u52a1\uff0c\u6240\u8c13\u7684\u201c\u5f00\u9501\u201d\u73af\u5883\u662f\u6307\u4e00\u4e2a\u865a\u62df\u7684\u201c\u9003\u8131\u7a7a\u95f4\u201d\uff0c\u7a7a\u95f4\u5185\u6709\u590d\u6742\u7684\u5c42\u7ea7\uff0c\u8981\u6c42\u667a\u80fd\u4f53\u5bf9\u62bd\u8c61\u3001\u6cdb\u5316\u7684\u56e0\u679c\u7ed3\u6784\u8fdb\u884c\u63a8\u7406\u3002<\/span>\u672c\u6587\u548c\u5148\u524d\u4e00\u7cfb\u5217\u4e0a\u4f73\u7684\u65e0\u6a21\u578b\u5f3a\u5316\u5b66\u4e60\u7b97\u6cd5\u8fdb\u884c\u4e86\u6bd4\u8f83\u3002<\/span>\u5f3a\u5316\u5b66\u4e60\u667a\u80fd\u4f53\u5728\u4e0d\u540c\u7684\u5c1d\u8bd5\u4e2d\u663e\u793a\u51fa\u8f83\u5dee\u7684\u53ef\u8fc1\u79fb\u77e5\u8bc6\u7684\u5b66\u4e60\u80fd\u529b\u3002<\/span>\u4f46\u662f\u672c\u6587\u63d0\u51fa\u7684\u6a21\u578b\u5c55\u73b0\u51fa\u8d8b\u8fd1\u4eba\u7c7b\u5b66\u4e60\u8005\u7684\u6027\u80fd\uff0c\u66f4\u91cd\u8981\u7684\u662f\uff0c\u5c55\u73b0\u51fa\u5728\u4e0d\u540c\u7684\u5c1d\u8bd5\u548c\u5b66\u4e60\u73af\u5883\u4e2d\u5c55\u73b0\u51fa\u53ef\u8fc1\u79fb\u7684\u884c\u4e3a\u3002<\/span>
<\/section>\n
<\/span>
<\/section>\n
\u3010\u5185\u5bb9\u6807\u7b7e\u3011Causal AI \u56e0\u679c\u673a\u5668\u5b66\u4e60<\/span><\/strong><\/em><\/section>\n

<\/section>\n
3. \u8fd1\u671f\u793e\u533a\u6d3b\u52a8<\/span><\/strong><\/section>\n

<\/section>\n
2020\u5e7411\u67081\u65e5\u665a8\u70b9\uff0c\u56e0\u679c\u79d1\u5b66\u4e0eCausal AI\u8bfb\u4e66\u4f1a\u7b2c\u516d\u671f\u2014\u2014\u201c\u6f5c\u7ed3\u679c\u6846\u67b6\u4e0b\u7684\u56e0\u679c\u6548\u5e94\u4f30\u8ba1\u201d\u5982\u671f\u8fdb\u884c\uff0c\u6d59\u6c5f\u5927\u5b66\u52a9\u7406\u6559\u6388\u51b5\u7428\u4f5c\u4e86\u7cbe\u5f69\u5206\u4eab\u3002<\/span><\/section>\n

<\/section>\n
\u56e0\u679c\u95ee\u9898\u5b58\u5728\u4e8e\u5f88\u591a\u9886\u57df\uff0c\u5982\u533b\u7597\u5065\u5eb7\u3001\u7ecf\u6d4e\u3001\u653f\u6cbb\u79d1\u5b66\u3001\u6570\u5b57\u8425\u9500\u7b49\u3002\u6bd4\u5982\u4e00\u79cd\u65b0\u7684\u836f\u7269\u662f\u5426\u6bd4\u65e7\u7684\u836f\u7269\u66f4\u6709\u7597\u6548\uff1f\u4e00\u4e2a\u65b0\u7684\u7b56\u7565\u662f\u5426\u80fd\u63d0\u5347\u9500\u91cf\uff1f\u4e00\u4e2a\u65b0\u7684\u653f\u7b56\u4f1a\u7ed9\u6c11\u4f17\u3001\u7ecf\u6d4e\u548c\u793e\u4f1a\u5e26\u6765\u591a\u5927\u7684\u5f71\u54cd\uff1f\u6240\u4ee5\u8fd9\u4e9b\u95ee\u9898\u90fd\u9700\u8981\u56e0\u679c\u63a8\u7406\u7684\u6280\u672f\u6765\u89e3\u51b3\u3002<\/span><\/section>\n
 <\/span><\/section>\n
\u4ec0\u4e48\u662f\u56e0\u679c\uff0c\u901a\u4fd7\u6765\u8bf4\uff0c\u56e0\u679c\u5728\u751f\u6d3b\u4e2d\u5f88\u666e\u904d\uff0c\u201c\u56e0\u201d\u5176\u5b9e\u5c31\u662f\u5f15\u8d77\u67d0\u79cd\u73b0\u8c61\u53d1\u751f\u7684\u539f\u56e0\uff0c\u800c\u201c\u679c\u201d\u5c31\u662f\u67d0\u79cd\u73b0\u8c61\u53d1\u751f\u540e\u4ea7\u751f\u7684\u7ed3\u679c\u3002\u4f46\u56e0\u679c\u6027\u5374\u5f88\u96be\u76f4\u63a5\u89c2\u6d4b\u5230\uff0c\u4e00\u822c\u5728\u89c2\u6d4b\u4e2d\u4f1a\u5f97\u5230\u4e8b\u4ef6\u4e4b\u95f4\u7684\u76f8\u5173\u6027\uff0c\u800c\u5728\u89c2\u5bdf\u6027\u7814\u7a76\u4e2d\u53d1\u5c55\u81ea\u52a8\u7edf\u8ba1\u65b9\u6cd5\u6765\u63a8\u65ad\u56e0\u679c\u6548\u5e94\u662f\u975e\u5e38\u56f0\u96be\u7684\u3002\u51b5\u7428\u63d0\u51fa\u4e86\u4e00\u4e9b\u5728\u73b0\u5b9e\u7684\u5927\u6570\u636e\u573a\u666f\u4e2d\u9762\u4e34\u56e0\u679c\u6548\u5e94\u4f30\u8ba1\u7684\u4e00\u4e9b\u6311\u6218\uff0c\u5305\u62ec\uff081\uff09\u9ad8\u7ef4\u548c\u566a\u58f0\u53d8\u91cf\uff0c\uff082\uff09\u53d8\u91cf\u4e4b\u95f4\u76f8\u4e92\u4f5c\u7528\u7684\u672a\u77e5\u6a21\u578b\u7ed3\u6784\uff0c\u548c\uff083\uff09\u8fde\u7eed\/\u590d\u6742\u5904\u7406\u53d8\u91cf\u3002\u4e3a\u4e86\u5e94\u5bf9\u8fd9\u4e9b\u6311\u6218\uff0c\u4ed6\u4eec\u63d0\u51fa\u4e86\u4ee5\u4e0b\u7684\u7b97\u6cd5\uff1a<\/span><\/section>\n

<\/span><\/section>\n
Data-Driven Variable Decomposition (D2VD) algorithm\uff1b<\/span><\/section>\n
Decomposed Representation Counterfactual Regression (DeR-CFR) model;<\/span><\/section>\n
Differentiated Confounder Balancing (DCB) algorithm;<\/span><\/section>\n
Generative Adversarial De-confounding (GAD) algorithm.<\/span><\/section>\n

<\/span><\/section>\n
\u76f8\u6bd4\u4e8e\u5f53\u524d\u5df2\u6709\u7684\u65b9\u6cd5\uff0c\u4ed6\u4eec\u63d0\u51fa\u7684\u8fd9\u4e9b\u7b97\u6cd5\u5728\u89c2\u5bdf\u6027\u7814\u7a76\u4e2d\u53ef\u4ee5\u5bf9\u56e0\u679c\u6548\u5e94\u4f5c\u51fa\u66f4\u7cbe\u786e\u548c\u7a33\u5065\u7684\u4f30\u8ba1\u3002<\/span>\u4e86\u89e3\u66f4\u591a\u8be6\u60c5\uff1a<\/span><\/section>\n
<\/span>https:\/\/mp.weixin.qq.com\/s\/Yx5wtwl8efBNQ_S-grKxbA<\/span><\/section>\n


<\/p>\n

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

\u7cfb\u5217\u8bfb\u4e66\u4f1a\u62a5\u540d<\/strong><\/span><\/p>\n

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

<\/strong><\/p>\n


<\/p>\n

\u5927\u6570\u636e\u65f6\u4ee3\u7684\u4e0b\u4e00\u573a\u53d8\u9769\u2014\u2014\u56e0\u679c\u9769\u547d\u6b63\u5728\u915d\u917f\u4e4b\u4e2d\uff0c\u901a\u8fc7\u878d\u5408\u56e0\u679c\u63a8\u7406\u548c\u673a\u5668\u5b66\u4e60\u800c\u6784\u5efa\u51fa\u6765\u7684Causal AI\u7cfb\u7edf\uff0c\u6709\u671b\u5960\u5b9a\u5f3a\u4eba\u5de5\u667a\u80fd\u7684\u57fa\u77f3\u3002<\/span><\/p>\n


<\/p>\n

\u96c6\u667a\u4ff1\u4e50\u90e8\u8054\u5408\u5317\u4eac\u667a\u6e90\u4eba\u5de5\u667a\u80fd\u7814\u7a76\u9662\uff0c\u9080\u8bf7\u4e86\u4e00\u6279\u5bf9\u56e0\u679c\u79d1\u5b66\u4e0eCausal AI\u611f\u5174\u8da3\u7684\u7814\u7a76\u8005\uff0c\u5f00\u5c55\u4e3a\u671f2-3\u4e2a\u6708\u7684\u7cfb\u5217\u7ebf\u4e0a\u8bfb\u4e66\u4f1a\uff0c\u7814\u8bfb\u7ecf\u5178\u548c\u524d\u6cbf\u8bba\u6587\uff0c\u5e76\u5c1d\u8bd5\u96c6\u4f53\u64b0\u5199\u4e00\u90e8\u4e66\u7c4d\u3002\u5982\u679c\u4f60\u4e5f\u4ece\u4e8b\u76f8\u5173\u7684\u7814\u7a76\u3001\u5e94\u7528\u5de5\u4f5c\uff0c\u6b22\u8fce\u62a5\u540d\uff0c\u53c2\u4e0e\u8bfb\u4e66\u4f1a\u7684\u8ba8\u8bba\uff01<\/span><\/p>\n

 <\/span><\/p>\n

\u65f6\u95f4\uff1a<\/span><\/strong>9\u670820\u65e5\u8d77\uff0c\u6bcf\u5468\u65e5\u665a19:00-21:00\uff0c\u6301\u7eed\u7ea62-3\u4e2a\u6708<\/span><\/section>\n
\u6a21\u5f0f\uff1a<\/span><\/strong>\u7ebf\u4e0a\u95ed\u95e8\u8bfb\u4e66\u4f1a\uff1b\u6536\u8d39-\u9000\u6b3e\u7684\u4fdd\u8bc1\u91d1\u6a21\u5f0f\uff1b\u8bfb\u4e66\u4f1a\u6210\u5458\u8ba4\u9886\u89e3\u8bfb\u8bba\u6587<\/span><\/section>\n
\u8d39\u7528\uff1a<\/span><\/strong>299\/\u4eba<\/span><\/section>\n

 <\/span><\/p>\n

\u4e86\u89e3\u8bfb\u4e66\u4f1a\u5177\u4f53\u89c4\u5219\u3001\u62a5\u540d\u8bfb\u4e66\u4f1a\u8bf7\u70b9\u51fb\u4e0b\u65b9\u6587\u7ae0\uff1a<\/span><\/p>\n

\u56e0\u679c\u79d1\u5b66\u4e0e Causal AI \u7cfb\u5217\u8bfb\u4e66\u4f1a | \u4f17\u5305\u51fa\u4e66<\/a><\/section>\n


<\/p>\n

\u76ee\u524d\u8bfb\u4e66\u4f1a\u5df2\u7ecf\u6709\u8d85\u8fc7130\u4f59\u4eba\u7684\u6d77\u5185\u5916\u9ad8\u6821\u79d1\u7814\u9662\u6240\u7684\u4e00\u7ebf\u79d1\u7814\u5de5\u4f5c\u8005\u4ee5\u53ca\u4e92\u8054\u7f51\u4e00\u7ebf\u4ece\u4e1a\u4eba\u5458\u53c2\u4e0e\uff0c\u5982\u679c\u4f60\u4e5f\u5bf9\u8fd9\u4e2a\u4e3b\u9898\u611f\u5174\u8da3\uff0c\u5c31\u5feb\u52a0\u5165\u6211\u4eec\u5427\uff01<\/span><\/p>\n

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

              <\/span><\/p>\n

\u56fe\u6ce8\uff1a\u9488\u5bf9\u8bfb\u4e66\u4f1a\u7684\u4e3b\u9898\uff0c\u7531\u53d1\u8d77\u4eba\u9f9a\u9e64\u626c\u8bbe\u7f6e\u597d\u4e86\u5185\u5bb9\u6846\u67b6\uff0c\u6bcf\u4e2a\u4e3b\u9898\u4e0b\u6709\u4e00\u4e2a\u8d1f\u8d23\u4eba\u6765\u8d1f\u8d23\u7ef4\u62a4\u7ec4\u7ec7\u76f8\u5173\u5185\u5bb9\uff0c\u76ee\u524d\u5df2\u7ecf\u5b9a\u597d\u7684\u5982\u56fe\u6240\u793a\uff0c\u6b22\u8fce\u5bf9\u4e3b\u9898\u611f\u5174\u8da3\u7684\u8054\u7cfb\u76f8\u5173\u8d1f\u8d23\u4eba\uff0c\u4ee5\u53ca\u6765\u8ba4\u9886\u76f8\u5173\u4e3b\u9898\u3002<\/em><\/span><\/p>\n

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


<\/p>\n

\u56e0\u679c\u79d1\u5b66\u793e\u533a\u7b80\u4ecb<\/strong>\uff1a\u5b83\u662f\u7531\u96c6\u667a\u4ff1\u4e50\u90e8<\/span>\u3001\u667a\u6e90\u793e\u533a\u5171\u540c\u63a8\u52a8\uff0c\u9762\u5411\u56e0\u679c\u79d1\u5b66\u9886\u57df\u7684\u5782\u76f4\u578b\u5b66\u672f\u8ba8\u8bba\u793e\u533a\uff0c\u76ee\u7684\u662f\u4fc3\u8fdb\u56e0\u679c\u79d1\u5b66\u4e13\u4e1a\u4eba\u58eb\u548c\u5174\u8da3\u7231\u597d\u8005\u4eec\u7684\u4ea4\u6d41\u548c\u5408\u4f5c\uff0c\u63a8\u8fdb\u56e0\u679c\u79d1\u5b66\u5b66\u672f\u3001\u4ea7\u4e1a\u751f\u6001\u7684\u5efa\u8bbe\u548c\u843d\u5730\uff0c\u5b55\u80b2\u65b0\u4e00\u4ee3\u56e0\u679c\u79d1\u5b66\u9886\u57df\u7684\u5b66\u672f\u4e13\u5bb6\u548c\u4ea7\u4e1a\u521b\u65b0\u8005\u3002<\/p>\n

   <\/span><\/p>\n

\u56e0\u679c\u79d1\u5b66\u793e\u533a\u6b22\u8fce\u60a8\u52a0\u5165\uff01<\/p>\n

 <\/span><\/p>\n

\u56e0\u679c\u79d1\u5b66\u793e\u533a\u613f\u666f<\/strong>\uff1a<\/em>\u56de\u7b54\u56e0\u679c\u95ee\u9898\u662f\u5404\u4e2a\u9886\u57df\u8feb\u5207\u7684\u9700\u6c42\uff0c\u5f53\u524d\u8bb8\u591a\u4e0d\u540c\u9886\u57df\uff08\u4f8b\u5982 AI \u548c\u7edf\u8ba1\u5b66\uff09\u90fd\u5728\u4f7f\u7528\u56e0\u679c\u63a8\u7406\uff0c\u4f46\u662f\u4ed6\u4eec\u6240\u4f7f\u7528\u7684\u8bed\u8a00\u548c\u6a21\u578b\u5404\u4e0d\u76f8\u540c\uff0c\u5bfc\u81f4\u8fd9\u4e9b\u9886\u57df\u79d1\u5b66\u5bb6\u4e4b\u95f4\u6c9f\u901a\u4ea4\u6d41\u56f0\u96be\u3002\u56e0\u6b64\u6211\u4eec\u5e0c\u671b\u6784\u5efa\u4e00\u4e2a\u793e\u533a\uff0c\u901a\u8fc7\u7ec4\u7ec7\u5927\u91cf\u5b66\u672f\u6d3b\u52a8\uff0c\u4f7f\u5f97\u79d1\u7814\u4eba\u5458\u80fd\u591f\u638c\u63e1\u7edf\u8ba1\u5b66\u7684\u6838\u5fc3\u601d\u60f3\uff0c\u719f\u7ec3\u4f7f\u7528\u5f53\u524d AI \u5404\u79cd\u6280\u672f\uff08\u4f8b\u5982 Pytorch\/Pyro \u642d\u5efa\u6df1\u5ea6\u6982\u7387\u6a21\u578b\uff09\uff0c\u4fc3\u8fdb\u5404\u4e2a\u9886\u57df\u7684\u7814\u7a76\u8005\u4ea4\u6d41\u548c\u601d\u7ef4\u78b0\u649e\uff0c\u4ece\u800c\u8ba9\u5404\u4e2a\u9886\u57df\u7684\u56e0\u679c\u63a8\u7406\u6709\u7740\u5171\u540c\u7684\u8303\u5f0f\uff0c\u751a\u81f3\u662f\u5171\u540c\u7684\u5de5\u7a0b\u5b9e\u8df5\u6807\u51c6\uff0c\u63a8\u52a8\u521a\u521a\u6210\u578b\u7684\u56e0\u679c\u79d1\u5b66\u5feb\u901f\u5411\u524d\u53d1\u5c55\u3002<\/strong>\u5177\u5907\u56e0\u679c\u63a8\u7406\u80fd\u529b\u7684\u4eba\u7c7b\u7d27\u5bc6\u534f\u4f5c\u521b\u9020\u4e86\u5f3a\u5927\u7684\u6587\u660e\uff0c\u6211\u4eec\u5e0c\u671b\u5728\u672a\u6765\u793e\u4f1a\u4e2d\uff0c\u56e0\u679c\u63a8\u7406\u878d\u5165\u5230\u6bcf\u4e2a\u5b66\u79d1\uff0c\u5c24\u5176\u662f\u7d27\u5bc6\u7ed3\u5408\u548c\u63d0\u5347 AI \uff0c\u671f\u5f85\u65e0\u6570\u5177\u5907\u6500\u767b\u56e0\u679c\u4e4b\u68af\u80fd\u529b\u7684 Agents (Causal AI) \u548c\u4eba\u7c7b\u4e00\u8d77\u534f\u4f5c\uff0c\u5171\u5efa\u4e0b\u4e00\u4ee3\u7684\u4eba\u7c7b\u6587\u660e\uff01<\/strong><\/p>\n

 <\/p>\n

\u5982\u679c\u60a8\u6709\u9002\u5f53\u7684\u6570\u5b66\u57fa\u7840\u548c\u4eba\u5de5\u667a\u80fd\u7814\u7a76\u7ecf\u9a8c\uff0c\u65e2\u6709\u79d1\u5b66\u5bb6\u7684\u597d\u5947\u5fc3\u4e5f\u6709\u5de5\u7a0b\u5e08\u601d\u7ef4\uff0c\u5e0c\u671b\u53c2\u4e0e\u5230\u201c\u56e0\u679c\u9769\u547d\u201d\u4e2d\uff0c\u6559\u4f1a\u673a\u5668\u56e0\u679c\u601d\u7ef4\uff0c\u4e3a\u56e0\u679c\u79d1\u5b66\u4f5c\u51fa\u8d21\u732e\uff0c\u8bf7\u52a0\u5165\u6211\u4eec\u5fae\u4fe1\u7fa4\uff1a<\/strong><\/span>\u626b\u63cf\u4e0b\u9762\u793e\u533a\u5c0f\u52a9\u624b\u4e8c\u7ef4\u7801\u52a0\u5165\uff08\u8bf7\u5907\u6ce8\u201c\u56e0\u679c\u79d1\u5b66\u201d\uff09<\/span><\/strong>\ud83d\udc47<\/strong><\/p>\n


<\/strong><\/p>\n

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

       <\/span><\/p>\n


<\/p>\n


<\/p>\n

\n


<\/p>\n


<\/p>\n

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

\u96c6\u667a\u4ff1\u4e50\u90e8QQ\u7fa4\uff5c877391004<\/span><\/strong><\/p>\n

\u5546\u52a1\u5408\u4f5c\u53ca\u6295\u7a3f\u8f6c\u8f7d\uff5cswarma@swarma.org
<\/span><\/strong><\/p>\n

\n

\u25c6 <\/span>\u25c6 <\/span>\u25c6<\/span><\/span><\/strong><\/strong><\/h1>\n<\/section>\n

\u641c\u7d22\u516c\u4f17\u53f7\uff1a\u96c6\u667a\u4ff1\u4e50\u90e8<\/span><\/strong><\/span><\/p>\n


<\/p>\n

\u52a0\u5165\u201c\u6ca1\u6709\u56f4\u5899\u7684\u7814\u7a76\u6240\u201d<\/span><\/strong><\/span><\/p>\n

\n

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

\u8ba9\u82f9\u679c\u7838\u5f97\u66f4\u731b\u70c8\u4e9b\u5427\uff01<\/span><\/p>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section><\/div>\n","protected":false},"excerpt":{"rendered":"

\u4e3a\u4e86\u5e2e\u52a9\u5927\u5bb6\u66f4\u597d\u5730\u4e86\u89e3\u56e0\u679c\u79d1\u5b66\u7684\u6700\u65b0\u79d1\u7814\u8fdb\u5c55\u548c\u8d44\u8baf\uff0c\u6211\u4eec\u56e0\u679c\u79d1\u5b66\u793e\u533a\u56e2\u961f\u672c\u5468\u6574\u7406\u4e86\u7b2c2\u671f\u300a\u56e0\u679c\u79d1\u5b66\u5468\u520a\u300b\uff0c\u4ece Causality, Causal Inference, Causal AI \u4e09\u4e2a\u7ef4\u5ea6\u9e1f\u77b0\uff0c\u63a8\u9001\u8fd1\u671f\u56e0\u679c\u79d1\u5b66\u503c\u5f97\u5173\u6ce8\u7684\u8bba\u6587\u548c\u8d44\u8baf\u4fe1\u606f\uff0c \u540c\u65f6\u6211\u4eec\u4e5f\u5c06\u5411\u5927\u5bb6\u4ecb\u7ecd\u793e\u533a\u6b63\u5728\u63a8\u8fdb\u7684\u6d3b\u52a8\u2014\u2014\u56e0\u679c\u79d1\u5b66\u4e0e…<\/p>\n","protected":false},"author":1,"featured_media":22035,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":[],"categories":[1],"tags":[],"special":[],"_links":{"self":[{"href":"https:\/\/swarma.org\/index.php?rest_route=\/wp\/v2\/posts\/22045"}],"collection":[{"href":"https:\/\/swarma.org\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/swarma.org\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/swarma.org\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/swarma.org\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=22045"}],"version-history":[{"count":0,"href":"https:\/\/swarma.org\/index.php?rest_route=\/wp\/v2\/posts\/22045\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/swarma.org\/index.php?rest_route=\/wp\/v2\/media\/22035"}],"wp:attachment":[{"href":"https:\/\/swarma.org\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=22045"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/swarma.org\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=22045"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/swarma.org\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=22045"},{"taxonomy":"special","embeddable":true,"href":"https:\/\/swarma.org\/index.php?rest_route=%2Fwp%2Fv2%2Fspecial&post=22045"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}