报 告 人】张诚，复旦大学信息管理与信息系统系教授
【报告时间】2022年3月31日14:30-16:30
【报告地点】伟伦508，校外老师同学可通过腾讯会议加入，会议 ID：116-641-682，会议密码：220331，会议链接：https://meeting.tencent.com/dm/XZ4fJ3hyMtGc。
【报告题目】条件独立假设下的因果关系识别方法 Approaching Causality Identification Under Conditional Independent Assumption
【摘 要】Identifying causality between variables in the presence of confounding factors is one major challenge in empirical studies. Although randomized controlled experiments in labs predominate in the assessment of causal effect, causality identification in large observational studies has received increasing attention in recent years with the widespread enthusiasm for big data applications. The study provides a general data-driven framework that balances human domain knowledge and endogeneity identification in developing theoretical models with empirical evidences. Furthermore, the new approach can be used to
identify not only dichotomous causal effects, but also continuous causal effects. The framework shows its high effectiveness through simulated and actual empirical data sets.
【个人简介】 张诚，复旦大学信息管理与信息系 统系教授，新加坡国立大学计算机学院博士。研 究方向为信息技术商业价值。在 MIS Quarterly、 Journal of Management Information Systems、 Journal on Computing 、 Production and Operations Management、Marketing Science、 Journal of Marketing 等期刊上发表多篇论文。
