YINS Distinguished Lecture Series
Research Scientist, Facebook Artificial Intelligence Research
Title: Going beyond average effects in experiments using machine learning
Host: Dan Spielman
Talk Summary: Most experimentalists focus analyses on the average treatment effect. However, the advent of larger data sets and new techniques from machine learning means that we can start learning more nuanced things. This talk will cover a series of principled ways to go beyond just looking at averages and instead looking for heterogeneity in effects. I will also touch on ways to combine different data sources to learn causal effects of interest. The talk will cover be focused on applications to questions from behavioral science (human cooperation, in group bias and risk preferences) as well as big online experiments (personalization on Facebook).
Alex Peysakhovich is a research scientist at Facebook Artificial Intelligence Research. He’s interested in understanding learning and decision-making (human and machine). He is also interested in causal inference and causal discovery, which he views as important components towards having truly smart decision-makers (again, either human or machine). He’s also a huge fan of data science for civic improvement and social good. His work has been published in a number of academic journals and he’s written popular press articles for WIRED and New York Times. Before Facebook he was a joint post-doc with David Rand at the Human Cooperation Lab (Yale Psychology) and Martin Nowak at the Program for Evolutionary Dynamics (Harvard Biology). He completed his PhD in Economics at Harvard University under the watchful eyes of Al Roth, Drew Fudenberg, David Laibson and Uma Karmarkar.