AI4OPT Seminar Series

Date: Thursday, September 21, 2023

Time: Noon – 1:00 pm

Location: 9th floor Atrium in Coda Building (756 W Peachtree St NW, Atlanta, GA 30308)

Live streamed Virtually at: https://gatech.zoom.us/j/99381428980

Speaker: Mohsen Bayati 

A General Method for Experimentation Under Network Interference 

Abstract: Randomized experiments serve as a powerful methodology for the data-driven evaluation of decisions or treatments. Yet, there are instances where the validity of experiments is compromised by network interference. This phenomenon emerges when the treatment of a unit, such as a patient or a user, not only influences its own outcome but also affects the outcomes of other interconnected units. This interference can bias the conventional methods employed to estimate treatment effects.

In the present study, we propose a new framework designed to account for network interference effects, even when these effects follow complex patterns that extend beyond the specialized models currently found in the literature. This framework is anchored by an analytical toolbox in high-dimensional settings, tailored for the design and inference of experiments in contexts where network interference is prevalent. Building on this framework, we introduce a practical algorithm to estimate the total treatment effect and empirically validate its performance across three distinct settings, each characterized by a unique interference structure (this is joint work with Sadegh Shirani).

Bio: Mohsen Bayati is a professor in the Operations, Information, and Technology group at the Graduate School of Business at Stanford University and an Amazon Scholar. His research focuses on data-driven decision-making and experiment design, particularly as they intersect with the healthcare and retail sectors. He utilizes tools from contextual multi-armed bandits, graphical models, and high-dimensional statistics. Mohsen received a BS in Mathematics from Sharif University of Technology and a PhD in Electrical Engineering from Stanford University. He was then a postdoctoral researcher at Microsoft Research and at Stanford University. His work was awarded the INFORMS Healthcare Applications Society's Best Paper (Pierskalla) Award in 2014 and 2016, the INFORMS Applied Probability Society's Best Paper Award in 2015, and the National Science Foundation CAREER Award.

Lunch will be served at the seminar. Please stop by 15 minutes before the seminar to pick up lunch.

To continue receiving all AI4OPT seminar announcements, please sign up for the mailing list at: https://lists.isye.gatech.edu/mailman/listinfo/ai4opt-seminars 

Videos of the past seminars can be seen on the AI4OPT webpage at:                                               https://www.ai4opt.org/seminars/past-seminars