Saved in:
Bibliographic Details
Main Authors: Xu, Zhitong, Wang, Haitao, Phillips, Jeff M, Zhe, Shandian
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2402.02746
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915190246735872
author Xu, Zhitong
Wang, Haitao
Phillips, Jeff M
Zhe, Shandian
author_facet Xu, Zhitong
Wang, Haitao
Phillips, Jeff M
Zhe, Shandian
contents A long-standing belief holds that Bayesian Optimization (BO) with standard Gaussian processes (GP) -- referred to as standard BO -- underperforms in high-dimensional optimization problems. While this belief seems plausible, it lacks both robust empirical evidence and theoretical justification. To address this gap, we present a systematic investigation. First, through a comprehensive evaluation across twelve benchmarks, we found that while the popular Square Exponential (SE) kernel often leads to poor performance, using Matérn kernels enables standard BO to consistently achieve top-tier results, frequently surpassing methods specifically designed for high-dimensional optimization. Second, our theoretical analysis reveals that the SE kernel's failure primarily stems from improper initialization of the length-scale parameters, which are commonly used in practice but can cause gradient vanishing in training. We provide a probabilistic bound to characterize this issue, showing that Matérn kernels are less susceptible and can robustly handle much higher dimensions. Third, we propose a simple robust initialization strategy that dramatically improves the performance of the SE kernel, bringing it close to state-of-the-art methods, without requiring additional priors or regularization. We prove another probabilistic bound that demonstrates how the gradient vanishing issue can be effectively mitigated with our method. Our findings advocate for a re-evaluation of standard BO's potential in high-dimensional settings.
format Preprint
id arxiv_https___arxiv_org_abs_2402_02746
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Standard Gaussian Process is All You Need for High-Dimensional Bayesian Optimization
Xu, Zhitong
Wang, Haitao
Phillips, Jeff M
Zhe, Shandian
Machine Learning
A long-standing belief holds that Bayesian Optimization (BO) with standard Gaussian processes (GP) -- referred to as standard BO -- underperforms in high-dimensional optimization problems. While this belief seems plausible, it lacks both robust empirical evidence and theoretical justification. To address this gap, we present a systematic investigation. First, through a comprehensive evaluation across twelve benchmarks, we found that while the popular Square Exponential (SE) kernel often leads to poor performance, using Matérn kernels enables standard BO to consistently achieve top-tier results, frequently surpassing methods specifically designed for high-dimensional optimization. Second, our theoretical analysis reveals that the SE kernel's failure primarily stems from improper initialization of the length-scale parameters, which are commonly used in practice but can cause gradient vanishing in training. We provide a probabilistic bound to characterize this issue, showing that Matérn kernels are less susceptible and can robustly handle much higher dimensions. Third, we propose a simple robust initialization strategy that dramatically improves the performance of the SE kernel, bringing it close to state-of-the-art methods, without requiring additional priors or regularization. We prove another probabilistic bound that demonstrates how the gradient vanishing issue can be effectively mitigated with our method. Our findings advocate for a re-evaluation of standard BO's potential in high-dimensional settings.
title Standard Gaussian Process is All You Need for High-Dimensional Bayesian Optimization
topic Machine Learning
url https://arxiv.org/abs/2402.02746