Decision-Focused Learning in Restless Multi-Armed Bandits with Application to Maternal and Child Care Domain

This paper studies restless multi-armed bandit (RMAB) problems with unknown arm transition dynamics but with known correlated arm features. The goal is to learn a model to predict transition dynamics given features, where the Whittle index policy solves the RMAB problems using predicted transitions. However, prior works often learn the model by maximizing the predictive accuracy instead of final RMAB solution quality, causing a mismatch between training and evaluation objectives. To address this shortcoming we propose a novel approach for decision-focused learning in RMAB that directly trains the predictive model to maximize the Whittle index solution quality. We present three key contributions: (i) we establish the differentiability of the Whittle index policy to support decision-focused learning; (ii) we significantly improve the scalability of previous decision-focused learning approaches in sequential problems; (iii) we apply our algorithm to the service call scheduling problem on a real-world maternal and child health domain. Our algorithm is the first for decision-focused learning in RMAB that scales to large-scale real-world problems. \end{abstract}