Scalarizing cost-effective multi-objective optimization algorithms made possible with kriging

Purpose – The purpose of this paper is threefold: to make explicitly clear the range of efficient multi-objective optimization algorithms which are available with kriging; to demonstrate a previously uninvestigated algorithm on an electromagnetic design problem; and to identify algorithms particularly worthy of investigation in this field. Design/methodology/approach – The paper concentrates exclusively on scalarizing multi-objective optimization algorithms. By reviewing the range of selection criteria based on kriging models for single-objective optimization along with the range of methods available for transforming a multi-objective optimization problem to a single-objective problem, the family of scalarizing multi-objective optimization algorithms is made explicitly clear. Findings – One of the proposed algorithms is demonstrated on the multi-objective design of an electron gun. It is able to identify efficiently an approximation to the Pareto-optimal front. Research limitations/implications – The algorithms proposed are applicable to unconstrained problems only. One future development is to incorporate constraint-handling techniques from single-objective optimization into the scalarizing algorithms. Originality/value – A family of algorithms, most of which have not been explored before in the literature, is proposed. Algorithms of particular potential (utilizing the most promising developments in single-objective optimization) are identified.