Near-Optimal Deployment Efficiency in Reward-Free Reinforcement Learning with Linear Function Approximation

TL;DR

This paper introduces a near-optimal algorithm for reward-free reinforcement learning with linear function approximation, achieving optimal deployment and sample complexities simultaneously, which is crucial for cost-sensitive real-world applications.

Contribution

It presents the first algorithm with optimal deployment efficiency and linear dependence in sample complexity for reward-free RL under linear MDPs.

Findings

Achieves $ ilde{O}(d^2H^5/\epsilon^2)$ trajectory complexity for $\\epsilon$-optimal policies.

Introduces exploration-preserving policy discretization and a generalized G-optimal experiment design.

Provides lower bounds for switching cost and batch complexity in low-adaptive RL.

Abstract

We study the problem of deployment efficient reinforcement learning (RL) with linear function approximation under the \emph{reward-free} exploration setting. This is a well-motivated problem because deploying new policies is costly in real-life RL applications. Under the linear MDP setting with feature dimension $d$ and planning horizon $H$, we propose a new algorithm that collects at most $\widetilde{O}(\frac{d^2H^5}{\epsilon^2})$ trajectories within $H$ deployments to identify $\epsilon$-optimal policy for any (possibly data-dependent) choice of reward functions. To the best of our knowledge, our approach is the first to achieve optimal deployment complexity and optimal $d$ dependence in sample complexity at the same time, even if the reward is known ahead of time. Our novel techniques include an exploration-preserving policy discretization and a generalized G-optimal experiment design, which could be of independent interest. Lastly, we analyze the related problem of regret minimization in low-adaptive RL and provide information-theoretic lower bounds for switching cost and batch complexity.