Title: A Bayesian Approach to Policy Recognition and State Representation Learning

Abstract: Learning from demonstration (LfD) is the process of building behavioral models of a task from demonstrations provided by an expert. These models can be used e.g. for system control by generalizing the expert demonstrations to previously unencountered situations. Most LfD methods, however, make strong assumptions about the expert behavior, e.g. they assume the existence of a deterministic optimal ground truth policy or require direct monitoring of the expert's controls, which limits their practical use as part of a general system identification framework. In this work, we consider the LfD problem in a more general setting where we allow for arbitrary stochastic expert policies, without reasoning about the quality of the demonstrations. In particular, we focus on the problem of policy recognition, which is to extract a system's latent control policy from observed system behavior. Following a Bayesian methodology allows us to consider various sources of uncertainty about the expert behavior, including the latent expert controls, to model the full posterior distribution of expert controllers. Further, we show that the same methodology can be applied in a nonparametric context to reason about the complexity of the state representation used by the expert and to learn task-appropriate partitionings of the system state space.