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

Methods for offline A/B testing and counterfactual learning are seeing rapidadoption in search and recommender systems, since they allow efficient reuse ofexisting log data. However, there are fundamental limits to using existing logdata alone, since the counterfactual estimators that are commonly used in thesemethods can have large bias and large variance when the logging policy is verydifferent from the target policy being evaluated. To overcome this limitation,we explore the question of how to design data-gathering policies that mosteffectively augment an existing dataset of bandit feedback with additionalobservations for both learning and evaluation. To this effect, this paperintroduces Minimum Variance Augmentation Logging (MVAL), a method forconstructing logging policies that minimize the variance of the downstreamevaluation or learning problem. We explore multiple approaches to computingMVAL policies efficiently, and find that they can be substantially moreeffective in decreasing the variance of an estimator than na\"ive approaches.

Variance-Optimal Augmentation Logging for Counterfactual Evaluation in Contextual Bandits