Learning Treatment Policies for Mobile Health

Computer Science/Statistics

This project involves investigating a recently proposed algorithm in reinforcement learning—Randomized Least-Squares Value Iteration—to develop an algorithm that can determine the optimal frequency at which to treat a user in mobile health. The interested student will conduct extensive coding, reading and running simulations, and will participate in writing a paper about how this method should be generalized to be useful in mobile health. Interested students must have a strong computational background.