Learning dynamic causal effects with potential processes

About the Talk:

I will detail a nonparametric model of time series causality, arguing that it is essential to make a “faithfulness” assumption to have any hope of drawing causal conclusions.  The framework will be related to existing methods. 

About the speaker:

Neil Shephard’s broad research interests are in econometrics, finance and statistics, with a particular focus on financial econometrics.   He has made significant advances in developing simulation based inference methods for online learning and has contributed methods to allow the mainstream use of high frequency financial data in economics. 

He joined the Harvard faculty in 2013 as Professor of Economics and of Statistics, holding the position equally between the Economics Department and the Statistics Departments.  He was chair of the Harvard University’s Department of Statistics from 2015 to 2022.   In 2018 he became the Frank B. Baird, Jr. Professor of Science, still working in the Economics and Statistics Departments.   

Professor Shephard is a fellow of the Econometric Society, the British Academy, the Society for Financial Econometrics and the International Association for Applied Econometrics.  Professor Shephard was a faculty member at the London School of Economics from 1988-1993 and Nuffield College, Oxford from 1991 to 2013.  He received his Ph.D. from the LSE in 1990.