PCIC 2021华为因果推理挑战赛

PCIC（the Pacific Causal Inference Conference，泛太平洋因果推断大会）由北京大学讲席教授、北京大学公共卫生学院生物统计系系主任、北京大学北京国际数学研究中心生物统计和信息研究事室主任周晓华组织发起，由北京大学、国际生物统计学会中国分会等机构主办、华为诺亚方舟实验室、集智俱乐部等机构协办。

PCIC是因果科学领域一年一度的盛会，PCIC 2021将于2021年9月11日至12日在北京以线上+线下相结合的方式举行。本大会聚集了海内外一大批因果科学相关领域的学术大拿汇聚一堂，共同探讨因果科学的新进展，探索新的科研方向，吸引了非常多优秀的科研工作者投身于该领域，推动了因果科学的研究进程。

会议期间会议组委会，北京大学和华为诺亚方舟实验室共同组织了此次因果竞赛。此次竞赛分为两个赛道。赛道1的目标是从连续时间事件型数据中发现因果关系。事件型数据在实际应用中会经常遇到，例如监测电信系统是否正常运行的报警记录。学习报警之间的因果关系可以用来识别根因报警，从而帮助快速定位系统的实际故障。赛道2则关注用户的偏好预测。实际中，每部电影通常都会有一定的描述性标签。参赛者基于关联数据（电影的标签）和因果数据（用户标明他喜欢某部电影的原因标签）建模，并纠偏有偏的推荐数据（用户选择偏置、前序模型偏置等），学习用户的兴趣标签。



竞赛的数据集为经过抽取脱敏的实际数据或者根据实际场景抽象出来的模拟数据。获奖团队获得现金奖励和证书，并邀请在大会上做技术报告。





时间安排




2021年06月08日：竞赛开放。

2021年06月15日：北京时间晚上11:59，初赛提交开放。

2021年07月15日：北京时间晚上11:59，报名和组队截止。

2021年08月10日: 北京时间晚上11:59，初赛结束，复赛开始。

2021年08月22日：北京时间晚上11:59，竞赛结束。

2021年09月01日：宣布获胜队伍。

2021年09月11日：PCIC 2021将邀请获奖团队进行技术报告。

最终的获奖队伍排名将由竞赛委员会根据竞赛成绩和提交方案的有效性评判。




竞赛奖励




每个赛道的奖金总额均为：$10,000（USD），其中

• Gold Prize: $5,000

• Silver Prize: $3,000

• Bronze Prize: $2,000

更多的信息请参加竞赛网站或者扫描二维码：
因果发现赛道：
https://competition.huaweicloud.com/information/1000041487/introduction
因果推理和推荐赛道：
https://competition.huaweicloud.com/information/1000041488/introduction




PCIC 2021 Huawei Causal Inference Competition


The goal of causal inference is to combine external knowledge and study design to draw a causal conclusion between variables. It has gained popularity in fields including statistics, biostatistics, biomedical science, computer science, economics, epidemiology, and various social sciences. To promote both research and application on real-world problems, Huawei Noah’s Ark Lab and Peking University jointly organize the competition focusing on causal inference and discovery at the 2021 Pacific Causal Inference Conference (PCIC 2021). The datasets used in the competition are either directly collected from or generated according to real scenarios.


The competition consists of two tracks. Track 1 aims to discover the causal graphs from event sequence data, which often occur in real applications such as social interactions amongst users or alarm records indicating whether a system operates properly. Obtaining such information is meaningful to better understand the relations among entities or to find root causes that can be more informative about the actual system fault. Track 2 focuses on the user preference prediction problem. In real-world movie recommendation systems, each movie is usually associated with descriptive tags, either by users or professionals. Instead of predicting the rating of a particular user-movie pair, participants are asked to estimate a user’s preference to a particular tag. The challenge comes from the existence of many biases, including user selection bias, popularity bias, etc.


The winners will be awarded cash prizes and winner certificates, as well as invited to give talks at PCIC 2021. This conference will focus on the latest development in causal inference and will take place between Sept. 11 and Sept. 12, 2021.

We look forward to your participation!




Timeline




  • June 8, 2021: Registration opens.

  • June 8, 2021: Competition opens.

  • June 15, 2021: Development phase starts at 11:59 pm Beijing Time.

  • July 15, 2021: Registration and team formation ends at 11:59 pm Beijing Time.

  • August 10, 2021: Development phase ends and test phase starts at 11:59 pm Beijing Time.

  • August 22, 2021: Competition ends at 11:59 pm Beijing Time

  • September 1, 2021: Winning teams are announced.

  • September 11, 2021: Winning teams will be invited to deliver a technical presentation at PCIC 2021.

The final rank of the winning teams will be determined by the competition committee based on the leaderboard scores from automated evaluation as well as the technical merits of the solutions.



Prize




Total Prize Amount (USD) for each track: $10,000

• Gold Prize: $5,000

• Silver Prize: $3,000

• Bronze Prize: $2,000

For more information, please visit the competition website or scan the QR codes:
Causal discovery track:
https://competition.huaweicloud.com/information/1000041487/introduction
Causal inference and recommendation track:
https://competition.huaweicloud.com/information/1000041488/introduction


因果科学第二季读书会报名中


因果推断与机器学习领域的结合已经吸引了越来越多来自学界业界的关注，为深入探讨、普及推广因果科学议题，智源社区携手集智俱乐部将举办第二季「因果科学与CausalAI读书会」。本期读书会着力于实操性、基础性，将带领大家精读因果科学方向两本非常受广泛认可的入门教材。

1. Pearl, Judea, Madelyn Glymour, and Nicholas P. Jewell. Causal inference in statistics: A primer. John Wiley & Sons, 2016.（本书中译版《统计因果推理入门（翻译版）》已由高等教育出版社出版）

2. Peters, Jonas, Dominik Janzing, and Bernhard Schölkopf. Elements of causal inference: foundations and learning algorithms. The MIT Press, 2017.

读书会每周将进行直播讨论，进行问题交流、重点概念分享、阅读概览和编程实践内容分析。非常适合有机器学习背景，希望深入学习因果科学基础知识和重要模型方法，寻求解决相关研究问题的朋友参加。

目前因果科学读书会系列，已经有接近400多位的海内外高校科研院所的一线科研工作者以及互联网一线从业人员参与，吸引了国内和国际上大部分的因果科学领域的专业科研人员，如果你也对这个主题感兴趣，想要深度地参与，就快加入我们吧！

详情请点击：
连接统计学、机器学习与自动推理的新兴交叉领域——因果科学读书会再起航


点击“阅读原文”，进入因果发现赛道