Integrating Domain Knowledge for handling Limited Data in Offline RL

TL;DR

This paper introduces a domain knowledge-based regularization method that enhances offline RL performance in limited data scenarios by reducing errors for rare and unseen states, showing significant empirical improvements.

Contribution

It proposes a novel regularization technique that adaptively refines domain knowledge to improve offline RL in data-scarce environments, addressing a key limitation of current algorithms.

Findings

Performance increased by at least 27% on standard datasets.

Regularization reduces errors for sparse and unobserved states.

Method effectively handles limited and partially omitted data.

Abstract

With the ability to learn from static datasets, Offline Reinforcement Learning (RL) emerges as a compelling avenue for real-world applications. However, state-of-the-art offline RL algorithms perform sub-optimally when confronted with limited data confined to specific regions within the state space. The performance degradation is attributed to the inability of offline RL algorithms to learn appropriate actions for rare or unseen observations. This paper proposes a novel domain knowledge-based regularization technique and adaptively refines the initial domain knowledge to considerably boost performance in limited data with partially omitted states. The key insight is that the regularization term mitigates erroneous actions for sparse samples and unobserved states covered by domain knowledge. Empirical evaluations on standard discrete environment datasets demonstrate a substantial average performance increase of at least 27% compared to existing offline RL algorithms operating on limited data.