Expensive Multiobjective Optimization by MOEA/D With Gaussian Process Model

TLDR

In expensive multiobjective optimization problems, batch evaluations of several objective functions can be performed simultaneously. The study aims to develop a method that generates multiple test points simultaneously for expensive multiobjective optimization. The proposed MOEA/D‑EGO algorithm decomposes the problem into single‑objective subproblems, builds a predictive model for each, uses MOEA/D to maximize expected improvement, and selects several test points per generation while reducing modeling overhead and improving prediction quality.

Abstract

In some expensive multiobjective optimization problems (MOPs), several function evaluations can be carried out in a batch way. Therefore, it is very desirable to develop methods which can generate multipler test points simultaneously. This paper proposes such a method, called MOEA/D-EGO, for dealing with expensive multiobjective optimization. MOEA/D-EGO decomposes an MOP in question into a number of single-objective optimization subproblems. A predictive model is built for each subproblem based on the points evaluated so far. Effort has been made to reduce the overhead for modeling and to improve the prediction quality. At each generation, MOEA/D is used for maximizing the expected improvement metric values of all the subproblems, and then several test points are selected for evaluation. Extensive experimental studies have been carried out to investigate the ability of the proposed algorithm.