Multi^3: Optimizing Multimodal Single-Objective Continuous Problems in the Multi-Objective Space by Means of Multiobjectivization

Aspar Pelin, Kerschke Pascal, Steinhoff Vera, Trautmann Heike, Grimme Christian


Zusammenfassung

In this work we examine the inner mechanisms of the recently developed sophisticated local search procedure SOMOGSA. This method solves multimodal single-objective continuous optimization problems by first expanding the problem with an additional objective (e.g., a sphere function) to the bi-objective space, and subsequently exploiting local structures and ridges of the resulting landscapes. Our study particularly focusses on the sensitivity of this multiobjectivization approach w.r.t. (i) the parametrization of the artificial second objective, as well as (ii) the position of the initial starting points in the search space.

As SOMOGSA is a modular framework for encapsulating local search, we integrate Gradient and Nelder-Mead local search (as optimizers in the respective module) and compare the performance of the resulting hybrid local search to their original single-objective counterparts. We show that the SOMOGSA framework can significantly boost local search by multiobjectivization. Combined with more sophisticated local search and metaheuristics this may help in solving highly multimodal optimization problems in future.

Schlüsselwörter
Multiobjective Optimization, Multimodalit



Publikationstyp
Forschungsartikel in Sammelband (Konferenz)

Begutachtet
Ja

Publikationsstatus
Veröffentlicht

Jahr
2021

Konferenz
11th International Conference on Evolutionary Multi-Criterion Optimization (EMO)

Konferenzort
Shenzhen, China

Buchtitel
Evolutionary Multi-Criterion Optimization: 11th International Conference, EMO 2021, Shenzhen, China, March 28–31, 2021, Proceedings

Herausgeber
Ishibuchi, H. et al.

Erste Seite
311

Letzte Seite
322

Verlag
Springer

Ort
Heidelberg, Berlin

Sprache
Englisch

DOI

Gesamter Text