Data-Efficient Off-Policy Policy Evaluation for Reinforcement Learning

TL;DR

This paper introduces a novel data-efficient off-policy evaluation method in reinforcement learning, significantly improving accuracy by combining model-based and importance sampling techniques to better predict policy performance from historical data.

Contribution

It extends the doubly robust estimator and proposes a new mixing approach, achieving lower mean squared error in policy evaluation compared to existing methods.

Findings

Estimates have orders of magnitude lower mean squared error.

The method makes more efficient use of available data.

Empirical results demonstrate superior performance.

Abstract

In this paper we present a new way of predicting the performance of a reinforcement learning policy given historical data that may have been generated by a different policy. The ability to evaluate a policy from historical data is important for applications where the deployment of a bad policy can be dangerous or costly. We show empirically that our algorithm produces estimates that often have orders of magnitude lower mean squared error than existing methods---it makes more efficient use of the available data. Our new estimator is based on two advances: an extension of the doubly robust estimator (Jiang and Li, 2015), and a new way to mix between model based estimates and importance sampling based estimates.