Reinforcement learning with human advice: a survey

TL;DR

This survey reviews various methods for incorporating human advice into reinforcement learning, categorizing advice types, interpretation techniques, and integration strategies to enhance learning efficiency and effectiveness.

Contribution

It provides a comprehensive taxonomy and overview of existing approaches for integrating human advice into reinforcement learning systems.

Findings

Taxonomy of advice types in reinforcement learning

Methods for interpreting ambiguous advice

Strategies for integrating advice into learning processes

Abstract

In this paper, we provide an overview of the existing methods for integrating human advice into a Reinforcement Learning process. We first propose a taxonomy of the different forms of advice that can be provided to a learning agent. We then describe the methods that can be used for interpreting advice when its meaning is not determined beforehand. Finally, we review different approaches for integrating advice into the learning process.