Improved Variance-Aware Confidence Sets for Linear Bandits and Linear Mixture MDP

TL;DR

This paper introduces variance-aware confidence sets for linear bandits and linear mixture MDPs, leading to regret bounds that adapt to unknown variances and improve existing results, especially in terms of dependence on horizon and episode count.

Contribution

The paper develops novel variance-aware confidence sets and technical tools, achieving the first regret bounds that scale only with variance and logarithmically with horizon, resolving open problems.

Findings

Regret bound for linear bandits scales with variance and dimension, independent of K.

Regret bound for linear mixture MDPs scales logarithmically with horizon H.

New technical tools include a variance recursion estimator and a convex potential lemma.

Abstract

This paper presents new \emph{variance-aware} confidence sets for linear bandits and linear mixture Markov Decision Processes (MDPs). With the new confidence sets, we obtain the follow regret bounds: For linear bandits, we obtain an $\tilde{O}(poly(d)\sqrt{1 + \sum_{k=1}^{K}\sigma_k^2})$ data-dependent regret bound, where $d$ is the feature dimension, $K$ is the number of rounds, and $\sigma_k^2$ is the \emph{unknown} variance of the reward at the $k$-th round. This is the first regret bound that only scales with the variance and the dimension but \emph{no explicit polynomial dependency on $K$}. When variances are small, this bound can be significantly smaller than the $\tilde{\Theta}\left(d\sqrt{K}\right)$ worst-case regret bound. For linear mixture MDPs, we obtain an $\tilde{O}(poly(d, \log H)\sqrt{K})$ regret bound, where $d$ is the number of base models, $K$ is the number of episodes, and $H$ is the planning horizon. This is the first regret bound that only scales \emph{logarithmically} with $H$ in the reinforcement learning with linear function approximation setting, thus \emph{exponentially improving} existing results, and resolving an open problem in \citep{zhou2020nearly}. We develop three technical ideas that may be of independent interest: 1) applications of the peeling technique to both the input norm and the variance magnitude, 2) a recursion-based estimator for the variance, and 3) a new convex potential lemma that generalizes the seminal elliptical potential lemma.