TGRL: An Algorithm for Teacher Guided Reinforcement Learning

TL;DR

TGRL introduces a principled, automatic method to balance teacher guidance and reward-based learning in reinforcement learning, improving performance without hyperparameter tuning.

Contribution

The paper proposes a novel, dynamic approach to balance teacher supervision and reward signals in reinforcement learning, eliminating the need for heuristics or extensive hyperparameter tuning.

Findings

TGRL outperforms strong baselines across various domains.

The method automatically adjusts the influence of teacher guidance.

It achieves better results without hyperparameter tuning.

Abstract

Learning from rewards (i.e., reinforcement learning or RL) and learning to imitate a teacher (i.e., teacher-student learning) are two established approaches for solving sequential decision-making problems. To combine the benefits of these different forms of learning, it is common to train a policy to maximize a combination of reinforcement and teacher-student learning objectives. However, without a principled method to balance these objectives, prior work used heuristics and problem-specific hyperparameter searches to balance the two objectives. We present a $\textit{principled}$ approach, along with an approximate implementation for $\textit{dynamically}$ and $\textit{automatically}$ balancing when to follow the teacher and when to use rewards. The main idea is to adjust the importance of teacher supervision by comparing the agent's performance to the counterfactual scenario of the agent learning without teacher supervision and only from rewards. If using teacher supervision improves performance, the importance of teacher supervision is increased and otherwise it is decreased. Our method, $\textit{Teacher Guided Reinforcement Learning}$ (TGRL), outperforms strong baselines across diverse domains without hyper-parameter tuning.