A comparative analysis of different infection strategies of Bacterial Memetic Algorithms

Abstract

Evolutionary methods and in particular Bacterial Memetic Algorithms are widely adopted means of population based metaheuristics, which have the ability to perform robust search on a discrete problem space. These methods are categorized as black-box search heuristics and tend to be quite good at finding generally good approximate solutions on certain problem domains such as the Traveling Salesman Problem. The good approximation ability is mainly credited to the bacterial infection operator, which helps to spread various suboptimal and partial solutions amongst the entire population. When gene transfer operations are omitted the heuristics is rendered to be a sole random sampling over the problem hyperspace. However there is a community dispute on the possible importance and effect of this operator on the search effectiveness in the case of optimization problems. Therefore in this paper the authors suggest multiple different infection strategies and perform a comparative analysis on their performance in the case of a real-life optimization scenario.