Benjamin Van Roy

activity

20062023

most citedDeep Exploration via Bootstrapped DQN

451 citations · 1.7k across the 73 of their papers we have counts for

collaborators

Showing 2019 · stat.MLShow all

3 papers · 2 filters

stat.ML2019★ 11 cited

Provably Efficient Reinforcement Learning with Aggregated States

Shi Dong, Benjamin Van Roy, Zhengyuan Zhou

We establish that an optimistic variant of Q-learning applied to a fixed-horizon episodic Markov decision process with an aggregated state representation incurs regret $\tilde{\mat…

stat.ML2019★ 13 cited

Information-Theoretic Confidence Bounds for Reinforcement Learning

Xiuyuan Lu, Benjamin Van Roy

We integrate information-theoretic concepts into the design and analysis of optimistic algorithms and Thompson sampling. By making a connection between information-theoretic quanti…

stat.ML2019★ 4 cited

On the Performance of Thompson Sampling on Logistic Bandits

Shi Dong, Tengyu Ma, Benjamin Van Roy

We study the logistic bandit, in which rewards are binary with success probability $exp (β a^{⊤} θ) / (1 + exp (β a^{⊤} θ))$ and actions $a$ and coefficients $θ$ are within the $d$ …