par Dubois-Lacoste, Jérémie 
;López-Ibáñez, Manuel ;Stützle, Thomas
Référence Lecture notes in computer science, 7245 LNCS, page (206-217)
Publication Publié, 2012
;López-Ibáñez, Manuel ;Stützle, Thomas Référence Lecture notes in computer science, 7245 LNCS, page (206-217)
Publication Publié, 2012
Article révisé par les pairs
| Résumé : | Pareto local search (PLS) is an extension of iterative improvement methods for multi-objective combinatorial optimization problems and an important part of several state-of-the-art multi-objective optimizers. PLS stops when all neighbors of the solutions in its solution archive are dominated. If terminated before completion, it may produce a poor approximation to the Pareto front. This paper proposes variants of PLS that improve its anytime behavior, that is, they aim to maximize the quality of the Pareto front at each time step. Experimental results on the bi-objective traveling salesman problem show a large improvement of the proposed anytime PLS algorithm over the classical one. © 2012 Springer-Verlag. |
