Pessimistic Model-based Offline Reinforcement Learning under Partial Coverage

TL;DR

This paper introduces CPPO, a model-based offline RL algorithm that achieves PAC guarantees under partial data coverage by leveraging a function class and a pessimism constraint, applicable to structured MDPs.

Contribution

The paper proposes CPPO, a novel offline RL algorithm with PAC guarantees under partial coverage, extending to structured MDPs like low-rank and factored models.

Findings

CPPO achieves PAC guarantees with partial data coverage.

The framework applies to low-rank and factored MDPs with structural assumptions.

Demonstrates effectiveness of pessimism-based constraints in offline RL.

Abstract

We study model-based offline Reinforcement Learning with general function approximation without a full coverage assumption on the offline data distribution. We present an algorithm named Constrained Pessimistic Policy Optimization (CPPO)which leverages a general function class and uses a constraint over the model class to encode pessimism. Under the assumption that the ground truth model belongs to our function class (i.e., realizability in the function class), CPPO has a PAC guarantee with offline data only providing partial coverage, i.e., it can learn a policy that competes against any policy that is covered by the offline data. We then demonstrate that this algorithmic framework can be applied to many specialized Markov Decision Processes where additional structural assumptions can further refine the concept of partial coverage. Two notable examples are: (1) low-rank MDP with representation learning where the partial coverage condition is defined using a relative condition number measured by the unknown ground truth feature representation; (2) factored MDP where the partial coverage condition is defined using density ratio based concentrability coefficients associated with individual factors.