2020/06/14 by S. Ashwin Renganathan, Renganathan, S. Ashwin, Jeffrey Larson +4
Computer Science · Decision Sciences · Mathematics · #Advanced Bandit Algorithms Research #Blind Source Separation Techniques #FOS: Computer and information sciences #FOS: Mathematics #Gaussian Processes and Bayesian Inference #Machine Learning (stat.ML) #Machine Learning and Algorithms #Optimization and Control (math.OC) #math.OC #stat.ML
paper · pdf · doi:10.48550/arxiv.2006.08037
24pages, 5 figures
openalex publication_date 2020/06/14 · arxiv created 2020/12/08 · arxiv updated 2020/12/09 · openalex created_date 2022/07/26 · openalex updated_date 2026/07/28
We propose a novel Bayesian method to solve the maximization of a time-dependent expensive-to-evaluate oracle. We are interested in the decision that maximizes the oracle at a finite time horizon, when relatively few noisy evaluations can be performed before the horizon. Our recursive, two-step lookahead expected payoff (r2LEY) acquisition function makes nonmyopic decisions at every stage by maximizing the estimated expected value of the oracle at the horizon. r2LEY circumvents the evaluation of the expensive multistep (more than two steps) lookahead acquisition function by recursively optimizing a two-step lookahead acquisition function at every stage; unbiased estimators of this latter function and its gradient are utilized for efficient optimization. r2LEY is shown to exhibit natural exploration properties far from the time horizon, enabling accurate emulation of the oracle, which is exploited in the final decision made at the horizon. To demonstrate the utility of r2LEY, we compare it with time-dependent extensions of popular myopic acquisition functions via both synthetic and real-world datasets.