Lightweight Interpolation-Based Surrogate Modelling for Multi-Objective Continuous Optimisation

Ciprian Zavoianu - Robert Gordon University, Aberdeen, UK

Abstract: We describe early results on how lightweight interpolation-based surrogate models can be used to reduce the prohibitive run-times observed when applying multi-objective evolutionary algorithms (MOEAs) on complex industrial design and manufacturing problems that require computationally-intensive performance assessment of tentative solutions (i.e., fitness evaluation).