|
Data Science Seminar Date: 23/12/2021 Day: Thursday Time: 6:30 pm - 7:30 pm(IST) 8:00 am - 9:00 am (EST) Nonparametric Bayesian Q-learning for adjusting partial compliance in Sequential Decision Making Indrabati Bhattacharya Department of Biostatistics and Computational Biology at University of Rochester, USA. 23-12-21 Abstract Q-learning is a well-known reinforcement learning approach for the estimation of optimal dynamic treatment regimes. Existing methods for estimation of dynamic treatment regimes are limited to intention-to-treat analyses--which estimate the effect of randomization to a particular treatment regime without considering the compliance behavior of patients. In this article, we propose a novel Bayesian nonparametric Q-learning approach based on stochastic decision rules for adjusting partial compliance. We consider the popular potential compliance framework, where some potential compliances are latent and need to be imputed. For each stage, we fit a locally weighted Dirichlet process mixture model for the potential outcome and compliance values. The key challenge is learning the joint distribution of the potential compliances, which we do using a Dirichlet process mixture model. Our approach provides two sets of decision rules: (1) conditional decision rules given the potential c ompliance values; and (2) marginal decision rules where the potential compliances are marginalized. Extensive simulation studies show the effectiveness of our method compared to intention-to-treat analyses. We apply our method on the Adaptive Treatment For Alcohol and Cocaine Dependence Study (ENGAGE), where the goal is to construct optimal treatment regimes to engage patients in therapy. Short bio: Indrabati Bhattacharya is a postdoctoral associate in the Department of Biostatistics and Computational Biology at the University of Rochester. Previously, she received her Ph.D. in Statistics from North Carolina State University. Her research interests include: Bayesian Asymptotics, Causal Inference, Machine Learning, Shape-restricted inference.
|
