Optimal Exploration–Exploitation in a Multi-armed Bandit Problem with Non-stationary Rewards

DOI10.1287/stsy.2019.0033zbMath1447.93371arXiv1405.3316OpenAlexW2962821829WikidataQ126855665 ScholiaQ126855665MaRDI QIDQ5113912

Yonatan Gur, Assaf J. Zeevi, Omar Besbes

Publication date: 18 June 2020

Published in: Stochastic Systems (Search for Journal in Brave)

Full work available at URL: https://arxiv.org/abs/1405.3316

zbMATH Keywords

minimax regretnonstationarymulti-armed banditdynamic oracleexploration/exploitationdynamic regret

Mathematics Subject Classification ID

Optimal stochastic control (93E20)

Related Items (7)

Robust sequential design for piecewise-stationary multi-armed bandit problem in the presence of outliers ⋮ Fully probabilistic design of strategies with estimator ⋮ Setting Reserve Prices in Second-Price Auctions with Unobserved Bids ⋮ Lipschitzness is all you need to tame off-policy generative adversarial imitation learning ⋮ Unnamed Item ⋮ Model-based preference quantification ⋮ Bayesian adversarial multi-node bandit for optimal smart grid protection against cyber attacks

Uses Software

• AdaBoost.MH

Cites Work

• Regret bounds for restless Markov bandits
• An analog of the minimax theorem for vector payoffs
• Asymptotically efficient adaptive allocation rules
• Arm-acquiring bandits
• A decision-theoretic generalization of on-line learning and an application to boosting
• Regret in the on-line decision problem
• Non-Stationary Stochastic Optimization
• On Upper-Confidence Bound Policies for Switching Bandit Problems
• Dynamic Assortment with Demand Learning for Seasonal Consumer Goods
• Restless Bandits, Linear Programming Relaxations, and a Primal-Dual Index Heuristic
• The Nonstochastic Multiarmed Bandit Problem
• Learning and Strategic Pricing
• Regret Analysis of Stochastic and Nonstochastic Multi-armed Bandit Problems
• Prediction, Learning, and Games
• Some aspects of the sequential design of experiments
• Finite-time analysis of the multiarmed bandit problem
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item

This page was built for publication: Optimal Exploration–Exploitation in a Multi-armed Bandit Problem with Non-stationary Rewards