Polyak-Ruppert-Averaged Q-Learning is Statistically Efficient

We study synchronous Q-learning with Polyak-Ruppert averaging (a.k.a., averaged Q-learning) in a $\gamma$-discounted MDP. We establish a functional central limit theorem (FCLT) for the averaged iteration $\bar{\boldsymbol{Q}}_T$ and show its standardized partial-sum process converges weakly to a rescaled Brownian motion. Furthermore, we show that $\bar{\boldsymbol{Q}}_T$ is actually a regular asymptotically linear (RAL) estimator for the optimal Q-value function $\boldsymbol{Q}^*$ with the most efficient influence function. This implies the averaged Q-learning iteration has the smallest asymptotic variance among all RAL estimators. In addition, we present a nonasymptotic analysis for the $\ell_{\infty}$ error $\mathbb{E}\|\bar{\boldsymbol{Q}}_T-\boldsymbol{Q}^*\|_{\infty}$, showing that for polynomial step sizes it matches the instance-dependent lower bound as well as the optimal minimax complexity lower bound. In short, our theoretical analysis shows that averaged Q-learning is statistically efficient.