Efficient Bayesian Optimization using Multiscale Graph Correlation

TL;DR

This paper introduces GP-MGC, a novel Bayesian optimization method that maximizes multiscale graph correlation to efficiently identify the global maximum, outperforming existing approaches in various applications.

Contribution

The paper presents a new Bayesian optimization approach using multiscale graph correlation, improving the efficiency and effectiveness over current state-of-the-art methods.

Findings

GP-MGC performs as well as or better than max-value entropy search and GP-UCB.

Demonstrated effectiveness on synthetic and real-world datasets.

Shows potential for more efficient optimization in costly black-box functions.

Abstract

Bayesian optimization is a powerful tool to optimize a black-box function, the evaluation of which is time-consuming or costly. In this paper, we propose a new approach to Bayesian optimization called GP-MGC, which maximizes multiscale graph correlation with respect to the global maximum to determine the next query point. We present our evaluation of GP-MGC in applications involving both synthetic benchmark functions and real-world datasets and demonstrate that GP-MGC performs as well as or even better than state-of-the-art methods such as max-value entropy search and GP-UCB.