Inverse reinforcement learning in continuous time and space

TL;DR

This paper introduces a data-driven inverse reinforcement learning method for continuous-time, continuous-space linear systems, enabling online estimation of an agent's cost function from input-output data.

Contribution

It develops a novel output-feedback inverse reinforcement learning approach using simultaneous state and parameter estimation for linear systems.

Findings

Successfully estimates cost functions from input-output data

Operates online in continuous-time and space settings

Achieves estimation up to a multiplicative constant

Abstract

This paper develops a data-driven inverse reinforcement learning technique for a class of linear systems to estimate the cost function of an agent online, using input-output measurements. A simultaneous state and parameter estimator is utilized to facilitate output-feedback inverse reinforcement learning, and cost function estimation is achieved up to multiplication by a constant.