Hybridization of genetic algorithm and local search in multiobjective function optimization: recommendation of GA then LS

Hybridization with local search (LS) is known to enhance the performance of genetic algorithms (GA) in single objective optimization and have also been studied in the multiobjective combinatorial optimization literature. In most such studies, LS is applied to the solutions of each generation of GA, which is the scheme called "GA with LS" herein. Another scheme, in which LS is applied to the solutions obtained with GA, has also been studied, which is called "GA then LS" herein. It seems there is no consensus in the literature as to which scheme is better, let alone the reasoning for it. The situation in the multiobjective function optimization literature is even more unclear since the number of such studies in the field has been small.This paper, assuming that objective functions are differentiable, reveals the reasons why GA is not suitable for obtaining solutions of high precision, thereby justifying hybridization of GA and LS. It also suggests that the hybridization scheme which maximally exploits both GA and LS is GA then LS. Experiments conducted on many benchmark problems verified our claims.

[1]  Marco Laumanns,et al.  Combining Convergence and Diversity in Evolutionary Multiobjective Optimization , 2002, Evolutionary Computation.

[2]  Masayuki Yamamura,et al.  Theoretical Analysis of Simplex Crossover for Real-Coded Genetic Algorithms , 2000, PPSN.

[3]  Marco Laumanns,et al.  Mutation Control and Convergence in Evolutionary Multi-Object Optimization , 2001 .

[4]  Kalyanmoy Deb,et al.  A fast and elitist multiobjective genetic algorithm: NSGA-II , 2002, IEEE Trans. Evol. Comput..

[5]  Kalyanmoy Deb,et al.  A Computationally Efficient Evolutionary Algorithm for Real-Parameter Optimization , 2002, Evolutionary Computation.

[6]  Marco Laumanns,et al.  SPEA2: Improving the Strength Pareto Evolutionary Algorithm For Multiobjective Optimization , 2002 .

[7]  Joshua D. Knowles,et al.  On metrics for comparing nondominated sets , 2002, Proceedings of the 2002 Congress on Evolutionary Computation. CEC'02 (Cat. No.02TH8600).

[8]  Clarisse Dhaenens,et al.  A Hybrid Evolutionary Approach for Multicriteria Optimization Problems: Application to the Flow Shop , 2001, EMO.

[9]  Lothar Thiele,et al.  Comparison of Multiobjective Evolutionary Algorithms: Empirical Results , 2000, Evolutionary Computation.

[10]  Martin Brown,et al.  Effective Use of Directional Information in Multi-objective Evolutionary Computation , 2003, GECCO.

[11]  Kalyanmoy Deb,et al.  Self-Adaptive Genetic Algorithms with Simulated Binary Crossover , 2001, Evolutionary Computation.

[12]  Hisao Ishibuchi,et al.  Some Issues on the Implementation of Local Search in Evolutionary Multiobjective Optimization , 2004, GECCO.

[13]  K. Deb,et al.  Hybrid methods for multi-objective evolutionary algorithms , 2002 .

[14]  Peter A. N. Bosman,et al.  Exploiting gradient information in numerical multi--objective evolutionary optimization , 2005, GECCO '05.

[15]  Joshua D. Knowles,et al.  Memetic Algorithms for Multiobjective Optimization: Issues, Methods and Prospects , 2004 .

[16]  Kalyanmoy Deb,et al.  A Fast Elitist Non-dominated Sorting Genetic Algorithm for Multi-objective Optimisation: NSGA-II , 2000, PPSN.

[17]  Isao Ono,et al.  A Real Coded Genetic Algorithm for Function Optimization Using Unimodal Normal Distributed Crossover , 1997, ICGA.

[18]  D. Corne,et al.  On Metrics for Comparing Non Dominated Sets , 2001 .

[19]  Marco Laumanns,et al.  Scalable multi-objective optimization test problems , 2002, Proceedings of the 2002 Congress on Evolutionary Computation. CEC'02 (Cat. No.02TH8600).