A study on the design issues of Memetic Algorithm

Over the recent years, there has been increasing research activities made on improving the efficacy of memetic algorithm (MA) for solving complex optimization problems. Particularly, these efforts have revealed the success of MA on a wide range of real world problems. MAs not only converge to high quality solutions, but also search more efficiently than their conventional counterparts. Despite the success and surge in interests on MAs, there is still plenty of scope for furthering our understanding on how and why synergy between population- based and individual learning searchers would lead to successful Memetic Algorithms. In this paper we outline several important design issues of Memetic Algorithms and present a systematic study on each. In particular, we conduct extensive experimental studies on the impact of each individual design issue and their relative impacts on memetic search performances by means of three commonly used synthetic problems. From the empirical studies obtained, we attempt to reveal the behaviors of several MA variants to enhance our understandings on MAs.

I. Introduction

Modern stochastic algorithms such as evolutionary algorithms (EA) draw inspiration from biological evolution. EAs, unlike conventional numerical optimization methods, produce new search points that do not use information about the local slope of the objective function and are thus not prone to stalling at local optima. Instead they involve a search from a “population” of solutions; making use of competitive selection, recombination and mutation operators to generate new solutions which are biased towards better regions of the search space. Further, they have shown considerable potentials for solving optimization problems that are characterized by non-convex, disjoint or noisy solution spaces. Modern stochastic optimizers, which have attracted a great deal of attention in recent years; include simulated annealing, tabu search, genetic algorithms, evolutionary programming, evolutionary strategies, differential evolution and many others [1], [2], [3] and [4]. These stochastic methods have been successfully applied to many real world optimization problems.