Bayesian Optimization under Stochastic Delayed Feedback

TL;DR

This paper develops Bayesian optimization algorithms that effectively handle stochastic delayed feedback, enabling more practical and efficient optimization in real-world scenarios with random feedback delays.

Contribution

It introduces new algorithms with sub-linear regret guarantees for Bayesian optimization under stochastic delays, extending to batch and contextual Gaussian process bandits.

Findings

Algorithms outperform existing methods in synthetic datasets

Effective handling of random feedback delays improves optimization efficiency

Validated on real-life datasets with positive results

Abstract

Bayesian optimization (BO) is a widely-used sequential method for zeroth-order optimization of complex and expensive-to-compute black-box functions. The existing BO methods assume that the function evaluation (feedback) is available to the learner immediately or after a fixed delay. Such assumptions may not be practical in many real-life problems like online recommendations, clinical trials, and hyperparameter tuning where feedback is available after a random delay. To benefit from the experimental parallelization in these problems, the learner needs to start new function evaluations without waiting for delayed feedback. In this paper, we consider the BO under stochastic delayed feedback problem. We propose algorithms with sub-linear regret guarantees that efficiently address the dilemma of selecting new function queries while waiting for randomly delayed feedback. Building on our results, we also make novel contributions to batch BO and contextual Gaussian process bandits. Experiments on synthetic and real-life datasets verify the performance of our algorithms.