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

This study addresses the problem of off-policy evaluation (OPE) fromdependent samples obtained via the bandit algorithm. The goal of OPE is toevaluate a new policy using historical data obtained from behavior policiesgenerated by the bandit algorithm. Because the bandit algorithm updates thepolicy based on past observations, the samples are not independent andidentically distributed (i.i.d.). However, several existing methods for OPE donot take this issue into account and are based on the assumption that samplesare i.i.d. In this study, we address this problem by constructing an estimatorfrom a standardized martingale difference sequence. To standardize thesequence, we consider using evaluation data or sample splitting with a two-stepestimation. This technique produces an estimator with asymptotic normalitywithout restricting a class of behavior policies. In an experiment, theproposed estimator performs better than existing methods, which assume that thebehavior policy converges to a time-invariant policy.

Confidence Interval for Off-Policy Evaluation from Dependent Samples via Bandit Algorithm: Approach from Standardized Martingales