The content describes the use of black-box optimization in finding optimal parameters for systems like Neural Networks or complex simulations. It introduces Polynomial-Model-Based Optimization (PMBO) as a novel black-box optimizer that fits a polynomial surrogate to the objective function to minimize it. Inspired by Bayesian optimization, PMBO updates its model iteratively using the Expected Improvement acquisition function, balancing exploitation and exploration and providing uncertainty estimates. PMBO is compared to other state-of-the-art algorithms and proves to be successful, outperforming them in some cases. Based on the results, PMBO is considered to be the preferred choice for solving black-box optimization tasks in various disciplines.
Optimizing Blackbox Objectives Using Polynomial Models (arXiv:2309.00663v1 [cs.LG])
