Advertisement

Natural Computing

, Volume 5, Issue 3, pp 229–255 | Cite as

The dynamics of the best individuals in co-evolution

  • Elena Popovici
  • Kenneth De Jong
Article

Abstract

There continues to be a growing interest in the use of co-evolutionary algorithms to solve difficult computational problems. However, their performance has varied widely from good to disappointing. The main reason for this is that co-evolutionary systems can display quite complex dynamics. Therefore, in order to efficiently use co-evolutionary algorithms for problem solving, one must have a good understanding of their dynamical behavior. To build such understanding, we have constructed a methodology for analyzing co-evolutionary dynamics based on trajectories of best-of-generation individuals. We applied this methodology to gain insights into how to tune certain algorithm parameters in order to improve performance.

Keywords

best-response co-evolution dynamics cooperative co-evolution optimization performance 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

Notes

Acknowledgements

We would like to thank: Gabriel Balan, for pointing out the “best-response” terminology as appropriate for our work; An anonymous reviewer of our paper (Popovici and De Jong, 2005a), for suggesting parameterizing the angle between the best-response curves; Jayshree Sarma, for proof reading the manuscript and providing useful comments that we incorporated for improving the paper.

References

  1. Alligood K, Sauer T and Yorke J (1996) Chaos: An Introduction to Dynamical Systems, SpringerGoogle Scholar
  2. Angeline PJ and Pollack JB (1994). ‘Coevolving high-level representations’. In: Langton C.G. (eds), Artificial Life III, Vol. XVII. Santa Fe, New Mexico, Addison-Wesley, pp. 55–71Google Scholar
  3. Bucci A and Pollack J (2002) Order-theoretic analysis of coevolution problems: coevolutionary statics. In: Langdon WB et al. (eds) Genetic and Evolutionary Computation Conference Workshop: Understanding Coevolution, Morgan KaufmannGoogle Scholar
  4. Cliff D and Miller GF (1994) Co-evolution of Pursuit and Evasion I: Biological and Game-theoretic Foundations. Technical Report, School of Cognitive and Computing Sciences, University of SussexGoogle Scholar
  5. Cliff D and Miller GF (1995) Tracking the red queen: measurements of adaptive progress in co-evolutionary simulations. European Conference on Artificial Life, pp. 200–218Google Scholar
  6. Ficici S, Melnik O and Pollack J (2000) A game-theoretic investigation of selection methods used in evolutionary algorithms. In: Zalzala eaA (ed.) Proc. of CEC2000, IEEE PressGoogle Scholar
  7. Ficici SG and Pollack JB (1998) Challenges in coevolutionary learning: arms-race dynamics, open-endedness, and mediocre stable states. In: Adami C, Belew RK, Kitano H and Taylor C (eds) Artificial Life VI: Proc. of the Sixth Int. Conf. on Artificial Life, The MIT Press, Cambridge, MA, pp. 238–247Google Scholar
  8. Hillis WD (1990) Co-evolving parasites improve simulated evolution as an optimization procedure. In: CNLS ’89: Proceedings of the Ninth Annual International Conference of the Center for Nonlinear Studies on Self-organizing, Collective, and Cooperative Phenomena in Natural and Artificial Computing Networks on Emergent Computation, North-Holland Publishing Co., pp. 228–234Google Scholar
  9. Hofbauer J and Sigmund K (1998) Evolutionary Games and Population Dynamics, Cambridge University PressGoogle Scholar
  10. Luke S and Wiegand RP (2003) When Coevolutionary Algorithms Exhibit Evolutionary Dynamics. In: Workshop Proceedings of the Genetic and Evolutionary Computation Conference (GECCO-2003), SpringerGoogle Scholar
  11. Pagie L and Mitchell M (2002). ‘A comparison of evolutionary and coevolutionary search’. International Journal of Computational Intelligence and Applications 2(1): 53–69CrossRefGoogle Scholar
  12. Panait L, Wiegand RP and Luke S (2004) A sensitivity analisys of a cooperative coevolutionary algorithm biased for optimization. In: Proceedings of the Genetic and Evolutionary Computation Conference – GECCO-2004Google Scholar
  13. Popovici E and De Jong K (2004) Understanding competitive co-evolutionary dynamics via fitness landscapes. In: Luke S (ed.) AAAI Fall Symposium on Artificial Multiagent Learning, AAAI PressGoogle Scholar
  14. Popovici E and De Jong K (2005a) A dynamical systems analysis of collaboration methods in cooperative co-evolution. In: AAAI Fall Symposium Series Co-evolution Workshop, AAAI PressGoogle Scholar
  15. Popovici E and De Jong K (2005b) relationships between internal and external metrics in co-evolution. In: Congress on Evolutionary Computation, pp. 2800–2807, IEEE PressGoogle Scholar
  16. Popovici E and De Jong K (2005c). Understanding cooperative co-evolutionary dynamics via simple fitness landscapes. In: Beyer H.-G. et al. (eds), Proceedings of the Genetic and Evolutionary Computation Conference, GECCO-2005. New York, ACM PressGoogle Scholar
  17. Potter M and De Jong K (1994) A cooperative coevolutionary approach to function optimization. In: Proceedings of the Third Conference on Parallel Problem Solving from Nature (PPSN III), Jerusalem, Israel, pp. 249–257, SpringerGoogle Scholar
  18. Rosin CD and Belew RK (1995) Methods for competitive co-evolution: finding opponents worth beating. In: Proceedings of the 6th International Conference on Genetic Algorithms, Morgan Kaufmann Publishers Inc., pp. 373–381Google Scholar
  19. Sims K. (1994). ‘Evolving 3D morphology and behaviour by competition’. In: Brooks R. and Maes P. (eds), Artificial Life IV Proceedings. MIT, Cambridge, MA, USA, MIT Press, pp. 28–39Google Scholar
  20. Wiegand RP (2004) An analysis of cooperative coevolutionary algorithms. Ph.D. thesis, George Mason University, Fairfax, VAGoogle Scholar
  21. Wiegand RP, Liles W and De Jong K (2002) Analyzing cooperative coevolution with evolutionary game theory. In: Proceedings of the Congress on Evolutionary Computation, IEEE Press, pp. 1600–1605Google Scholar
  22. Wiegand RP and Sarma J (2004). ‘Spatial embedding and loss of gradient in cooperative coevolutionary algorithms’. In: Yao X et al. (eds), 8th International Conference on Parallel Problem Solving from Nature - PPSN VIII. Birmingham, UK, Springer, pp. 912–921Google Scholar

Copyright information

© Springer Science+Business Media, Inc. 2006

Authors and Affiliations

  1. 1.George Mason UniversityFairfaxUSA

Personalised recommendations