Symmetry-Aware Bayesian Optimization via Max Kernels
Abstract
Bayesian Optimization (BO) is a powerful framework for optimizing noisy, expensive-to-evaluate black-box functions. When the objective exhibits invariances under a group action, exploiting these symmetries can substantially improve BO efficiency. While using maximum similarity across group orbits has long been considered in other domains, the fact that the max kernel is not positive semidefinite (PSD) has prevented its use in BO. In this work, we revisit this idea by considering a PSD projection of the max kernel. Compared to existing invariant (and non-invariant) kernels, we show it achieves significantly lower regret on both synthetic and real-world BO benchmarks, without increasing computational complexity.
Cite
Text
Bardou et al. "Symmetry-Aware Bayesian Optimization via Max Kernels." International Conference on Learning Representations, 2026.Markdown
[Bardou et al. "Symmetry-Aware Bayesian Optimization via Max Kernels." International Conference on Learning Representations, 2026.](https://mlanthology.org/iclr/2026/bardou2026iclr-symmetryaware/)BibTeX
@inproceedings{bardou2026iclr-symmetryaware,
title = {{Symmetry-Aware Bayesian Optimization via Max Kernels}},
author = {Bardou, Anthony and Gonon, Antoine and Ahadinia, Aryan and Thiran, Patrick},
booktitle = {International Conference on Learning Representations},
year = {2026},
url = {https://mlanthology.org/iclr/2026/bardou2026iclr-symmetryaware/}
}