Maximum Likelihood Constraint Inference for Inverse Reinforcement Learning
This addresses the challenge of more succinctly modeling expert behavior in control tasks for applications like robotics or autonomous systems, though it is incremental as it builds on the Maximum Entropy IRL framework.
The paper tackles the problem of Inverse Reinforcement Learning (IRL) by reformulating it to infer hard constraints (e.g., state, action, or feature constraints) that explain an expert agent's behavior, rather than just estimating a reward function, and demonstrates efficacy using simulated and human navigation data.
While most approaches to the problem of Inverse Reinforcement Learning (IRL) focus on estimating a reward function that best explains an expert agent's policy or demonstrated behavior on a control task, it is often the case that such behavior is more succinctly represented by a simple reward combined with a set of hard constraints. In this setting, the agent is attempting to maximize cumulative rewards subject to these given constraints on their behavior. We reformulate the problem of IRL on Markov Decision Processes (MDPs) such that, given a nominal model of the environment and a nominal reward function, we seek to estimate state, action, and feature constraints in the environment that motivate an agent's behavior. Our approach is based on the Maximum Entropy IRL framework, which allows us to reason about the likelihood of an expert agent's demonstrations given our knowledge of an MDP. Using our method, we can infer which constraints can be added to the MDP to most increase the likelihood of observing these demonstrations. We present an algorithm which iteratively infers the Maximum Likelihood Constraint to best explain observed behavior, and we evaluate its efficacy using both simulated behavior and recorded data of humans navigating around an obstacle.