Penalty Strategy in The Fitness Function of Grey Wolf Optimizer for Minimum Spanning Tree Problem

Abstract

The Grey Wolf Optimizer (GWO) is a relatively new population-based optimizer. Various optimization problems have been solved using GWO. This paper presents an experiment on how to apply GWO to solve the minimum spanning tree (MST) problem. MST is normally solved using revision strategy when formulating the fitness in other population-based algorithms. Another strategy, called Penalty strategy, is used in the experiment. Existing dataset for MST problem is tested using GWO. The experiment showed that implementation of penalty strategy in the fitness function of GWO can find a solution with almost 96% accuracy.