Guided Policy Optimization under Partial Observability

TL;DR

This paper introduces Guided Policy Optimization (GPO), a novel reinforcement learning framework for partially observable environments that leverages privileged information to improve policy learning, achieving near-optimal performance and outperforming existing methods.

Contribution

The paper proposes GPO, a new framework that co-trains a guider and a learner, effectively utilizing privileged information to enhance policy learning under partial observability.

Findings

GPO achieves near-optimal performance theoretically.

GPO significantly outperforms existing methods empirically.

GPO is effective in continuous control, noisy, and memory-based tasks.

Abstract

Reinforcement Learning (RL) in partially observable environments poses significant challenges due to the complexity of learning under uncertainty. While additional information, such as that available in simulations, can enhance training, effectively leveraging it remains an open problem. To address this, we introduce Guided Policy Optimization (GPO), a framework that co-trains a guider and a learner. The guider takes advantage of privileged information while ensuring alignment with the learner's policy that is primarily trained via imitation learning. We theoretically demonstrate that this learning scheme achieves optimality comparable to direct RL, thereby overcoming key limitations inherent in existing approaches. Empirical evaluations show strong performance of GPO across various tasks, including continuous control with partial observability and noise, and memory-based challenges, significantly outperforming existing methods.