Abstract
In test-based problems, commonly solved with competitive coevolution algorithms, candidate solutions (e.g., game strategies) are evaluated by interacting with tests (e.g., opponents). As the number of tests is typically large, it is expensive to calculate the exact value of objective function, and one has to elicit a useful training signal (search gradient) from the outcomes of a limited number of interactions between these coevolving entities. Averaging of interaction outcomes, typically used to that aim, ignores the fact that solutions often have to master different and unrelated skills, which form underlying objectives of the problem. We propose a method for on-line discovery of such objectives via heuristic compression of interaction outcomes. The compressed matrix implicitly defines derived search objectives that can be used by traditional multiobjective search techniques (NSGA-II in this study). When applied to the challenging variant of multi-choice Iterated Prisoner’s Dilemma problem, the proposed approach outperforms conventional two-population coevolution in a statistically significant way.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Axelrod, R.: The evolution of strategies in the iterated prisoner’s dilemma. The Dynamics of Norms, 1–16 (1987)
Bucci, A., Pollack, J.B., de Jong, E.: Automated extraction of problem structure. In: Deb, K., Tari, Z. (eds.) GECCO 2004, Part I. LNCS, vol. 3102, pp. 501–512. Springer, Heidelberg (2004)
Chong, S.Y., Tino, P., Ku, D.C., Xin, Y.: Improving Generalization Performance in Co-Evolutionary Learning. IEEE Transactions on Evolutionary Computation 16(1), 70–85 (2012)
Chong, S.Y., Yao, X.: Behavioral diversity, choices and noise in the iterated prisoner’s dilemma. IEEE Transactions on Evolutionary Computation 9(6), 540–551 (2005)
Darwen, P.J., Yao, X.: Why more choices cause less cooperation in iterated prisoner’s dilemma. In: Proceedings of the 2001 Congress on Evolutionary Computation, vol. 2, pp. 987–994. IEEE (2001)
de Jong, E.D., Bucci, A.: DECA: Dimension extracting coevolutionary algorithm. In: Cattolico, M.C., et al. (eds.) GECCO 2006: Proceedings of the 8th Annual Conference on Genetic and Evolutionary Computation, Seattle, Washington, USA, pp. 313–320. ACM Press (2006)
Deb, K., Pratap, A., Agarwal, S., Meyarivan, T.: A fast and elitist multiobjective genetic algorithm: NSGA-II. IEEE Transactions on Evolutionary Computation 6(2), 182–197 (2002)
Ficici, S.G., Pollack, J.B.: Pareto optimality in coevolutionary learning. In: Kelemen, J., Sosík, P. (eds.) ECAL 2001. LNCS (LNAI), vol. 2159, pp. 316–325. Springer, Heidelberg (2001)
Frean, M.: The evolution of degrees of cooperation. Journal of Theoretical Biology 182(4), 549–559 (1996)
Harrald, P.G., Fogel, D.B.: Evolving continuous behaviors in the iterated prisoner’s dilemma. Biosystems 37(1), 135–145 (1996)
Jaśkowski, W., Krawiec, K.: Formal analysis, hardness, and algorithms for extracting internal structure of test-based problems. Evolutionary Computation 19(4), 639–671 (2011)
Jaśkowski, W., Liskowski, P., Szubert, M., Krawiec, K.: Improving coevolution by random sampling. In: Blum, C. (ed.) GECCO 2013: Proceedings of the 15th Annual Conference on Genetic and Evolutionary Computation, Amsterdam, The Netherlands, pp. 1141–1148. ACM (2013)
Juillé, H., Pollack, J.B.: Coevolving the “ideal” trainer: Application to the discovery of cellular automata rules. University of Wisconsin, pp. 519–527. Morgan Kaufmann (1998)
Popovici, E., Bucci, A., Wiegand, R.P., de Jong, E.D.: Coevolutionary Principles. In: Handbook of Natural Computing. Springer (2011)
Poundstone, W.: Prisoner’s Dilemma: John von Neuman, Game Theory, and the Puzzle of the Bomb. Doubleday, New York (1992)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2014 Springer International Publishing Switzerland
About this paper
Cite this paper
Liskowski, P., Krawiec, K. (2014). Discovery of Implicit Objectives by Compression of Interaction Matrix in Test-Based Problems. In: Bartz-Beielstein, T., Branke, J., Filipič, B., Smith, J. (eds) Parallel Problem Solving from Nature – PPSN XIII. PPSN 2014. Lecture Notes in Computer Science, vol 8672. Springer, Cham. https://doi.org/10.1007/978-3-319-10762-2_60
Download citation
DOI: https://doi.org/10.1007/978-3-319-10762-2_60
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-10761-5
Online ISBN: 978-3-319-10762-2
eBook Packages: Computer ScienceComputer Science (R0)