A GA-VNS based algorithm for the multi-objective spanning tree problem

Abstract

The Multi-Objective Minimum Spanning Tree prob- lem (MOST ) has been shown to be NP -hard even with two criteria. In this study we propose a hybrid GA-VNS algorithm that exploits the advantages of both ”Non-dominated Sorting Genetic Algorithm” (NSGA-II) and ”Variable Neighborhood Search” (VNS) metaheuristics to find as good an approximation as possible to the Pareto front of MOST problem. Experimental studies provide the efficiency of the method which produces solutions as close as possible to the Pareto optimal front.