PSO and multi-funnel landscapes: how cooperation might limit exploration

Particle Swarm Optimization (PSO) is a population-based optimization method in which search points employ a cooperative strategy to move toward one another. In this paper we show that PSO appears to work well on "single-funnel" optimization functions. On more complex optimization problems, PSO tends to converge too quickly and then fail to make further progress. We contend that most benchmarks for PSO have classically been demonstrated on single-funnel functions. However, in practice, optimization tasks are more complex and possess higher problem dimensionality. We present empirical results that support our conjecture that PSO performs well on single-funnel functions but tends to stagnate on more complicated landscapes.

[1]  L. Darrell Whitley,et al.  The GENITOR Algorithm and Selection Pressure: Why Rank-Based Allocation of Reproductive Trials is Best , 1989, ICGA.

[2]  John N. Hooker,et al.  Testing heuristics: We have it all wrong , 1995, J. Heuristics.

[3]  Russell C. Eberhart,et al.  Comparison between Genetic Algorithms and Particle Swarm Optimization , 1998, Evolutionary Programming.

[4]  Russell C. Eberhart,et al.  Parameter Selection in Particle Swarm Optimization , 1998, Evolutionary Programming.

[5]  J. Doye,et al.  THE DOUBLE-FUNNEL ENERGY LANDSCAPE OF THE 38-ATOM LENNARD-JONES CLUSTER , 1998, cond-mat/9808265.

[6]  P. J. Angeline,et al.  Using selection to improve particle swarm optimization , 1998, 1998 IEEE International Conference on Evolutionary Computation Proceedings. IEEE World Congress on Computational Intelligence (Cat. No.98TH8360).

[7]  R. Eberhart,et al.  Empirical study of particle swarm optimization , 1999, Proceedings of the 1999 Congress on Evolutionary Computation-CEC99 (Cat. No. 99TH8406).

[8]  R. Eberhart,et al.  Comparing inertia weights and constriction factors in particle swarm optimization , 2000, Proceedings of the 2000 Congress on Evolutionary Computation. CEC00 (Cat. No.00TH8512).

[9]  Nikolaus Hansen,et al.  Completely Derandomized Self-Adaptation in Evolution Strategies , 2001, Evolutionary Computation.

[10]  Xiao-Feng Xie,et al.  Adaptive particle swarm optimization on individual level , 2002, 6th International Conference on Signal Processing, 2002..

[11]  Maurice Clerc,et al.  The particle swarm - explosion, stability, and convergence in a multidimensional complex space , 2002, IEEE Trans. Evol. Comput..

[12]  Jacques Riget,et al.  A Diversity-Guided Particle Swarm Optimizer - the ARPSO , 2002 .

[13]  James Kennedy,et al.  Bare bones particle swarms , 2003, Proceedings of the 2003 IEEE Swarm Intelligence Symposium. SIS'03 (Cat. No.03EX706).

[14]  Fabio Schoen,et al.  Global Optimization of Morse Clusters by Potential Energy Transformations , 2004, INFORMS J. Comput..

[15]  P. Pardalos.  RECENT ADVANCES AND TRENDS IN GLOBAL OPTIMIZATION : DETERMINISTIC AND STOCHASTIC METHODS , 2004 .

[16]  José Neves,et al.  The fully informed particle swarm: simpler, maybe better , 2004, IEEE Transactions on Evolutionary Computation.

[17]  Nikolaus Hansen,et al.  Evaluating the CMA Evolution Strategy on Multimodal Test Functions , 2004, PPSN.

[18]  Kevin D. Seppi,et al.  Exposing origin-seeking bias in PSO , 2005, GECCO '05.

[19]  Marco Locatelli,et al.  On the Multilevel Structure of Global Optimization Problems , 2005, Comput. Optim. Appl..

[20]  D. Wales.  Energy landscapes and properties of biomolecules , 2005, Physical biology.

[21]  J. Doye.  Physical Perspectives on the Global Optimization of Atomic Clusters , 2000, cond-mat/0007338.

[22]  Riccardo Poli,et al.  Particle swarm optimization , 1995, Swarm Intelligence.