English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

We study episodic reinforcement learning under unknown adversarialcorruptions in both the rewards and the transition probabilities of theunderlying system. We propose new algorithms which, compared to the existingresults in (Lykouris et al., 2020), achieve strictly better regret bounds interms of total corruptions for the tabular setting. To be specific, firstly,our regret bounds depend on more precise numerical values of total rewardscorruptions and transition corruptions, instead of only on the total number ofcorrupted episodes. Secondly, our regret bounds are the first of their kind inthe reinforcement learning setting to have the number of corruptions show upadditively with respect to $\min\{\sqrt{T}, \text{PolicyGapComplexity}\}$rather than multiplicatively. Our results follow from a general algorithmicframework that combines corruption-robust policy elimination meta-algorithms,and plug-in reward-free exploration sub-algorithms. Replacing themeta-algorithm or sub-algorithm may extend the framework to address othercorrupted settings with potentially more structure.

Improved Corruption Robust Algorithms for Episodic Reinforcement Learning