Finite-Time Analysis for the Knowledge-Gradient Policy

DOI10.1137/16M1073388zbMath1387.62029arXiv1606.04624OpenAlexW2963389017WikidataQ130050586 ScholiaQ130050586MaRDI QIDQ4610155

Yingfei Wang, Warren B. Powell

Publication date: 5 April 2018

Published in: SIAM Journal on Control and Optimization (Search for Journal in Brave)

Full work available at URL: https://arxiv.org/abs/1606.04624

zbMATH Keywords

stochastic controlranking and selectionsequential decision analysis

Mathematics Subject Classification ID

Bayesian problems; characterization of Bayes procedures (62C10) Learning and adaptive systems in artificial intelligence (68T05) Sequential statistical analysis (62L10) Statistical ranking and selection procedures (62F07)

Uses Software

• EGO
• BayesDA

Cites Work

• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Kullback-Leibler upper confidence bounds for optimal sequential allocation
• Optimal learning for sequential sampling with non-parametric beliefs
• Exploration-exploitation tradeoff using variance estimates in multi-armed bandits
• Asymptotically efficient adaptive allocation rules
• Efficient global optimization of expensive black-box functions
• Simulation budget allocation for further enhancing the efficiency of ordinal optimization
• Bayesian look ahead one-stage sampling allocations for selection of the best population
• Regret bounds for sleeping experts and bandits
• Global optimization of stochastic black-box systems via sequential kriging meta-models
• The Knowledge-Gradient Policy for Correlated Normal Beliefs
• The Knowledge-Gradient Algorithm for Sequencing Experiments in Drug Discovery
• On Upper-Confidence Bound Policies for Switching Bandit Problems
• Selecting a Selection Procedure
• A Knowledge-Gradient Policy for Sequential Information Collection
• An analysis of approximations for maximizing submodular set functions—I
• Sample mean based index policies by O(log n) regret for the multi-armed bandit problem
• The Data-Correcting Algorithm for the Minimization of Supermodular Functions
• Information-Theoretic Regret Bounds for Gaussian Process Optimization in the Bandit Setting
• Efficient Dynamic Simulation Allocation in Ordinal Optimization
• A Bayesian Approach to Some Best Population Problems
• Bandits With Heavy Tail
• Finite-time analysis of the multiarmed bandit problem