Constrained Bayesian Optimization Under Partial Observations: Balanced Improvements and Provable Convergence
This work addresses a domain-specific challenge in data-driven optimization for scenarios with partial observations, offering incremental improvements over existing methods.
The authors tackled partially observable constrained optimization problems (POCOPs) by developing a constrained Bayesian optimization method with improved acquisition functions and a Gaussian process surrogate model, achieving competitive results on synthetic and real-world problems.
The partially observable constrained optimization problems (POCOPs) impede data-driven optimization techniques since an infeasible solution of POCOPs can provide little information about the objective as well as the constraints. We endeavor to design an efficient and provable method for expensive POCOPs under the framework of constrained Bayesian optimization. Our method consists of two key components. Firstly, we present an improved design of the acquisition functions that introduces balanced exploration during optimization. We rigorously study the convergence properties of this design to demonstrate its effectiveness. Secondly, we propose a Gaussian process embedding different likelihoods as the surrogate model for a partially observable constraint. This model leads to a more accurate representation of the feasible regions compared to traditional classification-based models. Our proposed method is empirically studied on both synthetic and real-world problems. The results demonstrate the competitiveness of our method for solving POCOPs.