Is Separately Modeling Subpopulations Beneficial for Sequential Decision-Making?

Abstract

When applying decision models, we often estimate input parameters using data. In healthcare and some other applications, data are collected from a population of different entities, such as patients. Thus, one faces a modeling question of whether to estimate different models for subpopulations (called stratifying). The potential benefit of stratifying comes from the heterogeneity of subpopulations. For example, patients who progress faster than others require a separate model and a tailored treatment plan. In “Is Separately Modeling Subpopulations Beneficial for Sequential Decision-Making?,” Lee provides theoretical results and empirical methods for deciding whether to stratify subpopulations. The article also presents how to use its results to select the best stratification among many. Improving medical decisions by tailoring to each subpopulation is a building block of precision medicine, and thus, this work aligns closely with the precision medicine paradigm.