Learning What to Evaluate: Correlation-Aware Decoupling for Multiobjective Bayesian Optimization
Abstract
Multiobjective Bayesian optimization (MOBO) with Gaussian process (GP) surrogates is a sample efficient approach to solving multiobjective optimization problems. In MOBO, a Bayesian decision theoretic acquisition function guides the adaptive selection of new candidate inputs, on which objectives and constraints are evaluated to update the surrogate model sequentially. Existing approaches maintain independent GP models for the objectives and constraints, with new observations evaluating all objectives and constraints in a coupled fashion. However, the objectives and constraints often contain inherent correlations which, if exploited, can enable decoupled evaluations where only a subset of them are evaluated at each round. We present a new approach that leverages a multitask GP model to jointly learn all objectives and constraints, and propose a total correlation metric that enables identifying an optimal subset of objectives and constraints to be evaluated at every round, even under uniform evaluation costs. Theoretically, we show that our acquisition policy is asymptotically consistent despite decoupling and that our proposed decoupled subset selection rule maximizes the expected posterior entropy reduction about unevaluated tasks under mild conditions. Empirically, we show that our approach outperforms coupled and decoupled baselines in the state of the art.
Related
- Solving Multiobjective Linear Set Optimization Problems
- Optimality and Exact Moment-SDP Formulations for Robust Two-Stage Convex Polynomial Multiobjective Optimization
- Expected Hypervolume Maximization for Multiobjective Optimization under Uncertainties
- Calibrated α-Barycenters for Multiobjective f-Divergence Optimization
- Sample average approximation of multiobjective stochastic optimization problems with application in insurance
BibTeX
@misc{renganathan2026learning,
title = {{Learning What to Evaluate: Correlation-Aware Decoupling for Multiobjective Bayesian Optimization}},
author = {Ashwin Renganathan and Peter Bachman},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2609.32632},
eprint = {2609.32632},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.32632},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Learning What to Evaluate: Correlation-Aware Decoupling for Multiobjective Bayesian Optimization
AU - Ashwin Renganathan
AU - Peter Bachman
PY - 2026
DA - 2026/09/26
PB - arXiv
DO - 10.48550/arxiv.2609.32632
UR - https://arxiv.org/abs/2609.32632
AB - Multiobjective Bayesian optimization (MOBO) with Gaussian process (GP) surrogates is a sample efficient approach to solving multiobjective optimization problems. In MOBO, a Bayesian decision theoretic acquisition function guides the adaptive selection of new candidate inputs, on which objectives and constraints are evaluated to update the surrogate model sequentially. Existing approaches maintain independent GP models for the objectives and constraints, with new observations evaluating all objectives and constraints in a coupled fashion. However, the objectives and constraints often contain inherent correlations which, if exploited, can enable decoupled evaluations where only a subset of them are evaluated at each round. We present a new approach that leverages a multitask GP model to jointly learn all objectives and constraints, and propose a total correlation metric that enables identifying an optimal subset of objectives and constraints to be evaluated at every round, even under uniform evaluation costs. Theoretically, we show that our acquisition policy is asymptotically consistent despite decoupling and that our proposed decoupled subset selection rule maximizes the expected posterior entropy reduction about unevaluated tasks under mild conditions. Empirically, we show that our approach outperforms coupled and decoupled baselines in the state of the art.
ER -