Modelling the Population Distribution in Multi-objective Optimization by Generative Topographic Mapping

Under mild conditions, the Pareto set of a continuous multi-objective optimization problem exhibits certain regularity. We have recently advocated taking into consideration such regularity in designing multi-objective evolutionary algorithms. Following our previous work on using Local Principal Component Analysis for capturing the regularity, this paper presents a new approach for acquiring and using the regularity of the Pareto set in evolutionary algorithms. The approach is based on the Generative Topographic Mapping and can be regarded as an Estimation of Distribution Algorithm. It builds models of the distribution of promising solutions based on regularity patterns extracted from the previous search, and samples new solutions from the models thus built. The proposed algorithm has been compared with two other state-of-the-art algorithms, NSGA-II and SPEA2 on a set of test problems.

[1]  Tatsuya Okabe.  Evolutionary multi-objective optimization: on the distribution of offspring in parameter and fitness space , 2004 .

[2]  Satchidananda Dehuri,et al.  Evolutionary Algorithms for Multi-Criterion Optimization: A Survey , 2004 .

[3]  Dirk Thierens,et al.  The Naive MIDEA: A Baseline Multi-objective EA , 2005, EMO.

[4]  Nanda Kambhatla,et al.  Dimension Reduction by Local Principal Component Analysis , 1997, Neural Computation.

[5]  Bernhard Sendhoff,et al.  On Test Functions for Evolutionary Multi-objective Optimization , 2004, PPSN.

[6]  Qingfu Zhang,et al.  On stability of fixed points of limit models of univariate marginal distribution algorithm and factorized distribution algorithm , 2004, IEEE Transactions on Evolutionary Computation.

[7]  Marco Laumanns,et al.  A Tutorial on Evolutionary Multiobjective Optimization , 2004, Metaheuristics for Multiobjective Optimisation.

[8]  David E. Goldberg,et al.  Decomposable Problems, Niching, and Scalability of Multiobjective Estimation of Distribution Algorithms , 2005, ArXiv.

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

[10]  Jürgen Teich,et al.  Covering Pareto Sets by Multilevel Evolutionary Subdivision Techniques , 2003, EMO.

[11]  Marco Laumanns,et al.  Scalable Test Problems for Evolutionary Multiobjective Optimization , 2005, Evolutionary Multiobjective Optimization.

[12]  Christopher M. Bishop,et al.  GTM: The Generative Topographic Mapping , 1998, Neural Computation.

[13]  Yaochu Jin,et al.  Knowledge incorporation in evolutionary computation , 2005 .

[14]  Bernhard Sendhoff,et al.  Voronoi-based estimation of distribution algorithm for multi-objective optimization , 2004, Proceedings of the 2004 Congress on Evolutionary Computation (IEEE Cat. No.04TH8753).

[15]  Qingfu Zhang,et al.  On the convergence of a class of estimation of distribution algorithms , 2004, IEEE Transactions on Evolutionary Computation.

[16]  Qingfu Zhang,et al.  A model-based evolutionary algorithm for bi-objective optimization , 2005, 2005 IEEE Congress on Evolutionary Computation.

[17]  J. A. Lozano,et al.  Estimation of Distribution Algorithms: A New Tool for Evolutionary Computation , 2001 .

[18]  Yaochu Jin,et al.  Connectedness, regularity and the success of local search in evolutionary multi-objective optimization , 2003, The 2003 Congress on Evolutionary Computation, 2003. CEC '03..

[19]  Qingfu Zhang,et al.  Combining Model-based and Genetics-based Offspring Generation for Multi-objective Optimization Using a Convergence Criterion , 2006, 2006 IEEE International Conference on Evolutionary Computation.