Local Nonstationarity for Efficient Bayesian Optimization

TL;DR

This paper introduces a new nonstationary Gaussian process approach for Bayesian optimization, significantly improving performance on both stationary and nonstationary problems across various applications.

Contribution

It proposes a novel nonstationary strategy that enhances Bayesian optimization by relaxing the stationarity assumption of Gaussian processes.

Findings

Outperforms state-of-the-art Bayesian optimization methods

Effective on both stationary and nonstationary problems

Applicable across diverse application domains

Abstract

Bayesian optimization has shown to be a fundamental global optimization algorithm in many applications: ranging from automatic machine learning, robotics, reinforcement learning, experimental design, simulations, etc. The most popular and effective Bayesian optimization relies on a surrogate model in the form of a Gaussian process due to its flexibility to represent a prior over function. However, many algorithms and setups relies on the stationarity assumption of the Gaussian process. In this paper, we present a novel nonstationary strategy for Bayesian optimization that is able to outperform the state of the art in Bayesian optimization both in stationary and nonstationary problems.