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 Tools annual

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 Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Collaborative bandit learning, i.e., bandit algorithms that utilizecollaborative filtering techniques to improve sample efficiency in onlineinteractive recommendation, has attracted much research attention as it enjoysthe best of both worlds. However, all existing collaborative bandit learningsolutions impose a stationary assumption about the environment, i.e., both userpreferences and the dependency among users are assumed static over time.Unfortunately, this assumption hardly holds in practice due to users'ever-changing interests and dependence relations, which inevitably costs arecommender system sub-optimal performance in practice.  In this work, we develop a collaborative dynamic bandit solution to handle achanging environment for recommendation. We explicitly model the underlyingchanges in both user preferences and their dependency relation as a stochasticprocess. Individual user's preference is modeled by a mixture of globallyshared contextual bandit models with a Dirichlet Process prior. Collaborationamong users is thus achieved via Bayesian inference over the global banditmodels. Model selection and arm selection for each user are done via Thompsonsampling to balance exploitation and exploration. Our solution is proved tomaintain a standard $\tilde O(\sqrt{T})$ sublinear regret even in such achallenging environment. And extensive empirical evaluations on both syntheticand real-world datasets further confirmed the necessity of modeling a changingenvironment and our algorithm's practical advantages against severalstate-of-the-art online learning solutions.

When and Whom to Collaborate with in a Changing Environment: A Collaborative Dynamic Bandit Solution