CLAIMED: A CLAssification-Incorporated Minimum Energy Design to explore a multivariate response surface with feasibility constraints

TL;DR

This paper introduces CLAIMED, a novel method combining machine learning and experimental design to efficiently explore complex multivariate response surfaces with feasibility constraints, especially in high-dimensional physics simulations.

Contribution

It presents CLAIMED, a new approach that integrates classification and minimum energy design to identify multiple feasible regions in high-dimensional constrained optimization problems.

Findings

Effective identification of multiple feasible regions

Improved exploration of high-dimensional input spaces

Enhanced surrogate modeling accuracy

Abstract

Motivated by the problem of optimization of force-field systems in physics using large-scale computer simulations, we consider exploration of a deterministic complex multivariate response surface. The objective is to find input combinations that generate output close to some desired or "target" vector. In spite of reducing the problem to exploration of the input space with respect to a one-dimensional loss function, the search is nontrivial and challenging due to infeasible input combinations, high dimensionalities of the input and output space and multiple "desirable" regions in the input space and the difficulty of emulating the objective function well with a surrogate model. We propose an approach that is based on combining machine learning techniques with smart experimental design ideas to locate multiple good regions in the input space.