Wednesday, March 8, 2023 - 4:00pm to 4:30pm
Event Calendar Category
LIDS & Stats Tea
Building and Room Number
We propose a generalization of the synthetic controls and synthetic interventions methodology to incorporate network interference. We consider the estimation of unit-specific treatment effects from panel data where there are spillover effects across units and in the presence of unobserved confounding. Key to our approach is a novel latent factor model that takes into account network interference and generalizes the factor models typically used in panel data settings. We propose an estimator, "network synthetic interventions", and show that it consistently estimates the mean outcomes for a unit under an arbitrary sequence of treatments for itself and its neighborhood, given certain observation patterns hold in the data. We corroborate our theoretical findings with simulations.
Sarah Cen is a Ph.D. student at MIT co-advised by Devavrat Shah and Aleksander Madry. She is interested in responsible AI and causal inference. Her current projects look at the design of procedures for social media regulation, how strategic behavior affects the performance of recommender systems, and the effect of network interference on causal estimates.