An Enhanced MOGWW for the bi-objective Quadratic Assignment Problem
Everardo Gutiérrez-López, Carlos A. Brizuela · International Journal of Computational Intelligence Systems · 2011
This paper proposes an enhanced Multi-objective Go with the Winners (MOGWW) algorithm to solve multi-objective combinatorial optimization problems. The original MOGWW algorithm is equipped with the well known Pareto Local Search (PLS) procedure. In order to assess the performance of the hybridization, the non-dominated solutions it generates are compared with the ones generated by each of its components. The algorithms are applied to benchmark instances of the bi-objective Quadratic Assignment Problem. Experimental results show that the hybridized version outperforms both its components, i.e. the original MOGWW algorithm and a PLS variant.