Theoretically-Grounded Policy Advice from Multiple Teachers in Reinforcement Learning Settings with Applications to Negative Transfer

TL;DR

This paper introduces a formal framework for multiple teacher advice in reinforcement learning, providing theoretical guarantees and analyzing conditions under which negative transfer occurs, thereby advancing transfer learning understanding.

Contribution

It formalizes multi-teacher advice in RL, proposes an algorithm combining exploration and advice, and analyzes negative transfer with regret bounds.

Findings

Good teachers improve learning speed

Bad teachers can hinder learning

Negative transfer can be formally characterized

Abstract

Policy advice is a transfer learning method where a student agent is able to learn faster via advice from a teacher. However, both this and other reinforcement learning transfer methods have little theoretical analysis. This paper formally defines a setting where multiple teacher agents can provide advice to a student and introduces an algorithm to leverage both autonomous exploration and teacher's advice. Our regret bounds justify the intuition that good teachers help while bad teachers hurt. Using our formalization, we are also able to quantify, for the first time, when negative transfer can occur within such a reinforcement learning setting.