Abstract: In the stochastic K-armed bandit problem, an agent or learner is given a set of K probability distributions (informally, `arms').

Abstract: Many stochastic optimization and empirical dynamic programming algorithms have been proposed in the literature that approximates certain deterministic learning algorithms.

Abstract: Can computer modeling methods be used to design music instruments with improved intonation or response? Is that $6M Stradivarius violin really worth it? Why do clarinet and saxophone reeds wear out?

Abstract:  Discrete-time linear systems are a ubiquitous modeling tool across the sciences. We study a reachability problem related to discrete-time linear systems which we call ``Affine Subspace Reachability Problem''.