Select a calendar:
Filter February Events by Event Type:
SUNMONTUEWEDTHUFRISAT
Events for February 24, 2025
-
CS Colloquium: Angela Zhou (USC / Marshall School of Business) - Robust Fitted-Q-Evaluation and Iteration under Sequentially Exogenous Unobserved Confounders
Mon, Feb 24, 2025 @ 10:00 AM - 11:00 AM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Angela Zhou, USC / Marshall School of Business
Talk Title: Robust Fitted-Q-Evaluation and Iteration under Sequentially Exogenous Unobserved Confounders
Abstract: Offline causal decision making and reinforcement learning is important in domains such as medicine, economics, and e-commerce where online experimentation is costly, dangerous or unethical, and where the true model is unknown. However, most methods assume all covariates used in the behavior policy's action decisions are observed. Though this assumption, sequential ignorability/unconfoundedness, likely does not hold in observational data, most of the data that accounts for selection into treatment may be observed, motivating sensitivity analysis. We study robust policy evaluation and policy optimization in the presence of sequentially-exogenous unobserved confounders under a sensitivity model. We consider the single-timestep and the sequential setting. For the sequential setting, we propose and analyze orthogonalized robust fitted-Q-iteration that uses closed-form solutions of the robust Bellman operator to derive a loss minimization problem for the robust Q function, and adds a bias-correction to quantile estimation. Our algorithm enjoys the computational ease of fitted-Q-iteration and statistical improvements (reduced dependence on quantile estimation error) from orthogonalization. We provide sample complexity bounds, insights, and show effectiveness both in simulations and on real-world longitudinal healthcare data of treating sepsis. In particular, our model of sequential unobserved confounders yields an online Markov decision process, rather than partially observed Markov decision process: we illustrate how this can enable warm-starting optimistic reinforcement learning algorithms with valid robust bounds from observational data.
This lecture satisfies requirements for CSCI 591: Research Colloquium
Biography: Angela Zhou is an Assistant Professor in Data Sciences and Operations at the University of Southern California, Marshall School of Business. She received her PhD from Cornell ORIE and completed a postdoctoral fellowship at UC Berkeley / the Simons Institute. She works on data-driven decision making, including the interface of causal inference and machine learning, (offline) reinforcement learning, and equitable social prediction in consequential domains. She was a program co-chair for ACM EAAMO 2022 (a new conference on Equity and Access in Algorithms, Mechanisms and Optimization). Her research interests are in statistical machine learning for data-driven sequential decision making under uncertainty, causal inference, and the interplay of statistics and optimization. Her work has received oral-equivalent or featured designations at machine learning venues (Neurips, TMLR) and has won the INFORMS Data Mining Best Student Paper award, while she has received various designations as a Rising Star in AI, Data Science, and AI Fairness.
Host: CS Department
Location: Olin Hall of Engineering (OHE) - 132
Audiences: Everyone Is Invited
Contact: CS Faculty Affairs
