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 propose Banker Online Mirror Descent (Banker-OMD), a novel frameworkgeneralizing the classical Online Mirror Descent (OMD) technique in the onlinelearning literature. The Banker-OMD framework almost completely decouplesfeedback delay handling and the task-specific OMD algorithm design, thusfacilitating the design of new algorithms capable of efficiently and robustlyhandling feedback delays. Specifically, it offers a general methodology forachieving $\widetilde{\mathcal O}(\sqrt{T} + \sqrt{D})$-style regret bounds inonline bandit learning tasks with delayed feedback, where $T$ is the number ofrounds and $D$ is the total feedback delay. We demonstrate the power of\texttt{Banker-OMD} by applications to two important bandit learning scenarioswith delayed feedback, including delayed scale-free adversarial Multi-ArmedBandits (MAB) and delayed adversarial linear bandits. \texttt{Banker-OMD} leadsto the first delayed scale-free adversarial MAB algorithm achieving$\widetilde{\mathcal O}(\sqrt{K}L(\sqrt T+\sqrt D))$ regret and the firstdelayed adversarial linear bandit algorithm achieving $\widetilde{\mathcalO}(\text{poly}(n)(\sqrt{T} + \sqrt{D}))$ regret. As a corollary, the firstapplication also implies $\widetilde{\mathcal O}(\sqrt{KT}L)$ regret fornon-delayed scale-free adversarial MABs, which is the first to match the$\Omega(\sqrt{KT}L)$ lower bound up to logarithmic factors and can be ofindependent interest.

Banker Online Mirror Descent: A Universal Approach for Delayed Online Bandit Learning