报告题目：Counterfactual No-Harm Criterion: Individual Risk and Trustworthy Policy Learning

报告人：吴鹏 副教授 北京工商大学

时间：2023年10月11日（周三）15:00—16:00

地点：腾讯会议（会议 ID：993-537-556）

报告摘要:

Individual risk and trustworthy policy learning has significant importance in making reliable and harmless treatment decisions for individuals. Previous policy learning approaches aim at the well-being of subgroups by maximizing the utility function, however, individual-level counterfactual no-harm criterion has rarely been discussed. In this paper, we first formalize the counterfactual no-harm criterion from a principal stratification perspective. Next, we propose a novel upper bound for the fraction negatively affected by the policy and show the consistency and asymptotic normality of the estimator. Based on the estimators for the policy utility and harm upper bounds, we further propose a policy learning approach that satisfies the counterfactual no-harm criterion, and proves its consistency to the optimal policy reward for parametric and non-parametric policy classes, respectively.

报告人简介:

吴鹏，现为bat365官网登录入口，副教授。中国现场统计研究会因果推断分会理事，北京生物医学统计与数据管理研究会理事，ACM会员。吴鹏于2020年获得北京师范大学统计学博士学位，2020-2022年在北京大学北京国际数学研究中心从事博士后研究。研究方向包括因果推断，缺失数据，因果推荐系统，医疗决策，机器学习等。在著名杂志和国际计算机顶级会议发表论文20余篇。