Search papers, labs, and topics across Lattice.
This paper introduces a multi-source framework for constrained Bayesian Optimization (BO) that leverages auxiliary data to enhance exploration in environments with unknown constraints. By integrating multiple information sources, the method captures inter-source correlations and optimally balances evaluation costs with information gain. Experimental results demonstrate that this approach significantly outperforms traditional methods, particularly in the early stages of exploration, leading to more efficient identification of feasible and optimal solutions.
Auxiliary data can dramatically enhance constrained Bayesian Optimization, even when weakly correlated, leading to superior exploration and solution identification.
Bayesian Optimisation (BO) under unknown constraints is particularly challenging when feasible regions are small. In such settings, existing methods that typically rely solely on evaluations of the true objective and constraints struggle to efficiently explore the design space. However, many real-world applications offer auxiliary data sources (e.g. surrogate models or simplified simulations) that can support early exploration. Despite this potential, their integration into constrained BO remains largely unexplored. We propose a general multi-source framework that extends constrained Max-value Entropy Search, capturing inter-source correlation while balancing evaluation cost and information gain. Experiments on both synthetic and physics-based benchmarks show that our method efficiently identifies feasible and optimal solutions, even when auxiliary data are only weakly correlated. The proposed approach consistently outperforms existing methods, particularly in early-stage exploration.
