Online Program Home
  My Program

All Times EDT

Abstract Details

Activity Number: 325 - Machine Learning Methods for Better-Informed Decision-Making in Heath Care
Type: Invited
Date/Time: Thursday, August 12, 2021 : 10:00 AM to 11:50 AM
Sponsor: Health Policy Statistics Section
Abstract #316740
Title: Balance Competing Outcomes in Dynamic Treatment Regime Estimation
Author(s): Michael Kosorok and Siyeon Kim*
Companies: University of North Carolina at Chapel Hill and University of North Carolina
Keywords: Machine learning; Dynamic treatment regimes; Multiple outcomes; Inverse reinforcement learning

In this presentation, we discuss some new machine learning tools which allow researchers to balance competing outcomes in dynamic treatment regime estimation. One approach uses patient preference scores based on information gathered from patients through appropriate instruments. The second approach uses inverse reinforcement learning methods applied to observational data to learn how experts prioritize competing outcomes, even if they are not conscious of their prioritization and even if they do it with error.

Authors who are presenting talks have a * after their name.

Back to the full JSM 2021 program