Using BART to Perform Pareto Optimization and Quantify its Uncertainties

Techniques to reduce the energy burden of an industrial ecosystem often require solving a multiobjective optimization problem. However, collecting experimental data can be either expensive or time-consuming. In such cases, statistical methods helpful. This article proposes Pareto Front (PF) and Set (PS) estimation using Bayesian Additive Regression Trees (BART), which is nonparametric model whose assumptions are typically less restrictive than popular alternatives, as Gaussian Processes (GPs). These allow BART handle scenarios (e.g., high-dimensional input spaces, nonsmooth responses, large datasets) that GPs find difficult. The performance our BART-based method compared GP-based analytic test functions, demonstrating convincing advantages. Finally, methodology applied motivating engineering Supplementary materials, include theorem proof, algorithms, R code, for this available online.