arXiv:2311.15238v2 Announce Type: replace
Abstract: The exploration-exploitation dilemma has been a central challenge in reinforcement learning (RL) with complex model classes. In this paper, we propose a new algorithm, Monotonic Q-Learning with Upper Confidence Bound (MQL-UCB) for RL with general function approximation. Our key algorithmic design includes (1) a general deterministic policy-switching strategy that achieves low switching cost, (2) a monotonic value function structure with carefully controlled function class complexity, and (3) a variance-weighted regression scheme that exploits historical trajectories with high data efficiency. MQL-UCB achieves minimax optimal regret of $tilde{O}(dsqrt{HK})$ when $K$ is sufficiently large and near-optimal policy switching cost of $tilde{O}(dH)$, with $d$ being the eluder dimension of the function class, $H$ being the planning horizon, and $K$ being the number of episodes.
Our work sheds light on designing provably sample-efficient and deployment-efficient Q-learning with nonlinear function approximation.