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

Robust reinforcement learning (RRL) aims at seeking a robust policy tooptimize the worst case performance over an uncertainty set of Markov decisionprocesses (MDPs). This set contains some perturbed MDPs from a nominal MDP(N-MDP) that generate samples for training, which reflects some potentialmismatches between training (i.e., N-MDP) and true environments. In this paperwe present an elaborated uncertainty set by excluding some implausible MDPsfrom the existing sets. Under this uncertainty set, we develop a sample-basedRRL algorithm (named ARQ-Learning) for tabular setting and characterize itsfinite-time error bound. Also, it is proved that ARQ-Learning converges as fastas the standard Q-Learning and robust Q-Learning while ensuring betterrobustness. We introduce an additional pessimistic agent which can tackle themajor bottleneck for the extension of ARQ-Learning into the cases with largeror continuous state spaces. Incorporating this idea into RL algorithms, wepropose double-agent algorithms for model-free RRL. Via experiments, wedemonstrate the effectiveness of the proposed algorithms.

On Practical Robust Reinforcement Learning: Practical Uncertainty Set and Double-Agent Algorithm