arXiv：cs.LG· Rikuto Matsumoto, Masanori Ishikura, Masayuki Karasuyama·· 4 小时前AI 评分30
面向约束多保真多目标贝叶斯优化的统一信息论方法
A Unified Information-Theoretic Approach to Constrained Multi-Fidelity Multi-Objective Bayesian Optimization
AI 导读
研究提出一种统一信息论框架，用于约束条件下的多保真多目标贝叶斯优化，通过观测对最高保真可行 Pareto 前沿的信息增益来度量查询效用。由于该互信息难以计算，作者利用对 Pareto 一致区域的欠截断与过截断混合近似推导出变分下界，并借助多保真代理模型将信息传播至任意保真度，从而得到无需单独启发式的成本感知采集函数。合成、基准及真实问题实验显示该方法在多种目标、约束与保真设置下均有效。
正文
Abstract:Bayesian optimization often involves multiple objectives, constraints, and fidelity levels. We address the challenge of jointly selecting where and at which fidelity to evaluate to identify the highest-fidelity feasible Pareto frontier in this combined setting. From a unified information-theoretic perspective, we measure query utility by the information gain about this frontier, provided by an observation. Since this mutual information is intractable, we derive a variational lower bound using a mixture of under- and over-truncated approximations to the Pareto-consistent region. Multi-fidelity surrogate models propagate the information to arbitrary fidelities, yielding a cost-aware acquisition function without separate heuristics for fidelity selection or constraint handling. Experiments on synthetic, benchmark, and real-world problems demonstrate effectiveness across diverse objective, constraint, and fidelity settings.
| Subjects: | Machine Learning (cs.LG) |
| Cite as: | arXiv:2610.10174 [cs.LG] |
| (or arXiv:2610.10174v1 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2610.10174 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Masayuki Karasuyama [view email]
[v1]
Wed, 7 Oct 2026 14:47:10 UTC (1,667 KB)
来源：arXiv：cs.LG · arxiv.org