Advertisement

Soft Computing

, Volume 17, Issue 3, pp 503–518 | Cite as

Evolving optimum populations with XCS classifier systems

XCS with code fragmented action
  • Muhammad IqbalEmail author
  • Will N. Browne
  • Mengjie Zhang
Original Paper

Abstract

The main goal of the research direction is to extract building blocks of knowledge from a problem domain. Once extracted successfully, these building blocks are to be used in learning more complex problems of the domain, in an effort to produce a scalable learning classifier system (LCS). However, whilst current LCS (and other evolutionary computation techniques) discover good rules, they also create sub-optimum rules. Therefore, it is difficult to separate good building blocks of information from others without extensive post-processing. In order to provide richness in the LCS alphabet, code fragments similar to tree expressions in genetic programming are adopted. The accuracy-based XCS concept is used as it aims to produce maximally general and accurate classifiers, albeit the rule base requires condensation (compaction) to remove spurious classifiers. Serendipitously, this work on scalability of LCS produces compact rule sets that can be easily converted to the optimum population. The main contribution of this work is the ability to clearly separate the optimum rules from others without the need for expensive post-processing for the first time in LCS. This paper identifies that consistency of action in rich alphabets guides LCS to optimum rule sets.

Keywords

Learning classifier systems XCS Optimal populations Scalability Code fragments Action consistency 

References

  1. Acampora G, Cadenas JM, Loia V, Ballester EM (2011) A multi-agent memetic system for human-based knowledge selection. IEEE Trans Syst Man Cybern A Systems Humans 41(5):946–960 Google Scholar
  2. Ahluwalia M, Bull L (1999) A genetic programming based classifier system. In: Proceedings of the genetic and evolutionary computation conference, pp 11–18Google Scholar
  3. Alfaro-Cid E, Merelo JJ, de Vega FF, Esparcia-Alcázar AI, Sharman K (2010) Bloat control operators and diversity in genetic programming: a comparative study. Evol Comput 18(2):305–332CrossRefGoogle Scholar
  4. Altenberg L (1995) The schema theorem and Price’s theorem. In: Foundations of genetic algorithms, pp 23–49Google Scholar
  5. Banzhaf W, Nordin P, Keller RE, Francone FD (1998) Genetic programming—an introduction: on the automatic evolution of computer programs and its applications. Morgan Kaufmann, BurlingtonGoogle Scholar
  6. Bernad-Mansilla E, Garrell-Guiu JM (2003) Accuracy-based learning classifier systems: models, analysis and applications to classification tasks. Evol Comput 11(3):209–238CrossRefGoogle Scholar
  7. Beyer HG (1997) An alternative explanation for the manner in which genetic algorithms operate. BioSystems 41:1–15CrossRefGoogle Scholar
  8. Booker LB, Goldberg DE, Holland JH (1989) Classifier systems and genetic algorithms. Artif Intell 40(1-3):235–282CrossRefGoogle Scholar
  9. Burjorjee KM (2008) The fundamental problem with the building block hypothesisGoogle Scholar
  10. Butz MV (2000) XCSJava 1.0: an implementation of the XCS classifier system in Java. Technical Report 2000027, Illinois Genetic Algorithms LaboratoryGoogle Scholar
  11. Butz MV (2007) Combining gradient-based with evolutionary online learning: an introduction to learning classifier systems. In: Proceedings of the seventh international conference on hybrid intelligent systems, pp 12–17Google Scholar
  12. Butz MV, Kovacs T, Lanzi PL, Wilson SW (2001) How XCS evolves accurate classifiers. Technical Report 2001008, Illinois Genetic Algorithms LaboratoryGoogle Scholar
  13. Butz MV, Kovacs T, Lanzi PL, Wilson SW (2004) Toward a theory of generalization and learning in XCS. IEEE Trans Evol Comput 8(1):28–46CrossRefGoogle Scholar
  14. Butz MV, Pelikan M, Llorá X, Goldberg DE (2006) Automated global structure extraction for effective local building block processing in XCS. Evol Comput 14(3):345–380CrossRefGoogle Scholar
  15. Butz MV, Wilson SW (2002) An algorithmic description of XCS. Soft Comput A Fusion Found Methodol Appl 6(3-4):144–153zbMATHGoogle Scholar
  16. Drugowitsch J (2008) Design and analysis of learning classifier systems: a probabilistic approach. Springer BerlinGoogle Scholar
  17. Eiben AE, Smith JE (2003) Introduction to evolutionary computing, 1st edn. Natural Computing Series. Springer, BerlinGoogle Scholar
  18. Goldberg DE (1989) Genetic algorithms in search, optimization and machine learning. Addison Wesley, BostonGoogle Scholar
  19. Holland JH (1975) Adaptation in natural and artificial systems. University of Michigan Press, Ann ArborGoogle Scholar
  20. Holland JH (1986) Escaping brittleness: the possibilities of general-purpose learning algorithms applied to parallel rule-based systems. In: Machine learning: an artificial intelligence approach, vol II. Morgan Kaufmann, Burlington, pp 593–623Google Scholar
  21. Ioannides C, Browne WN (2007) Investigating scaling of an abstracted LCS utilising ternary and S-expression alphabets. In: Proceedings of the genetic and evolutionary computation conference, pp 2759–2764Google Scholar
  22. Jong KAD (2006) Evolutionary computation: a unified approach. MIT Press, CambridgeGoogle Scholar
  23. Kinzett D, Johnston M, Zhang M (2009) Numerical simplification for bloat control and analysis of building blocks in genetic programming. Evol Intell 2(4):151–168CrossRefGoogle Scholar
  24. Kovacs T (1996) Evolving optimal populations with XCS classifier systems. Technical Report CSR-96-17 and CSRP-9617, University of Birmingham, UKGoogle Scholar
  25. Koza JR (1992) Genetic programming: on the programming of computers by means of natural selection. MIT Press, CambridgeGoogle Scholar
  26. Koza JR, Poli R (2005) Genetic programming. In: Search methodologies: introductory tutorials in optimization and decision support techniques, chap. 5. Springer, Berlin, pp 127–164Google Scholar
  27. Lanzi PL (1999) Extending the representation of classifier conditions Part I: from binary to messy coding. In Proceedings of the genetic and evolutionary computation conference, pp 337–344Google Scholar
  28. Lanzi PL, Loiacono D (2007) Classifier systems that compute action mappings. In: Proceedings of the genetic and evolutionary computation conference, pp 1822–1829Google Scholar
  29. Lanzi PL, Loiacono D, Wilson SW, Goldberg DE (2005) XCS with computed prediction for the learning of Boolean functions. Technical Report 2005007, Illinois Genetic Algorithms LaboratoryGoogle Scholar
  30. Lanzi PL, Loiacono D, Wilson SW, Goldberg DE (2007) Generalization in the XCSF classifier system: analysis, improvement, and extension. Evol Comput 15(2):133–168CrossRefGoogle Scholar
  31. Lanzi PL, Perrucci A (1999) Extending the representation of classifier conditions Part II: from messy coding to S-expressions. In: Proceedings of the genetic and evolutionary computation conference, pp 345–352Google Scholar
  32. Lanzi PL, Stolzmann W, Wilson SW (2000) Learning classifier systems: from foundations to applications. Springer, BerlinGoogle Scholar
  33. Loiacono D, Marelli A, Lanzi P (2007) Support vector machines for computing action mappings in learning classifier systems. In: Proceedings of the congress on evolutionary computation, pp 2141–2148Google Scholar
  34. Luke S, Panait L (2006) A comparison of bloat control methods for genetic programming. Evol Comput 14(3):309–344CrossRefGoogle Scholar
  35. Mhlenbein H, Paaß G (1996) From recombination of genes to the estimation of distributions I. Binary parameters. In: Parallel Problem Solving from Nature, pp 178–187Google Scholar
  36. Orriols-Puig A, Bernadó-Mansilla E (2006) A further look at UCS classifier system. In: Proceedings of the ninth international workshop on learning classifier systems. Springer, BerlinGoogle Scholar
  37. Pelikan M, Goldberg DE, Lobo FG (2002) A survey of optimization by building and using probabilistic models. Comput Optim Appl 21(1):5–20MathSciNetzbMATHCrossRefGoogle Scholar
  38. Poli R (2000) Why the schema theorem is correct also in the presence of stochastic effects. In: Proceedings of the congress on evolutionary computation, pp 487–492Google Scholar
  39. Poli R, Langdon WB (1998) Schema theory for genetic programming with one-point crossover and point mutation. Evol Comput 6:231–252CrossRefGoogle Scholar
  40. Poli R, Langdon WB, McPhee NF (2008) A field guide to genetic programming. Lulu Enterprises, UK LtdGoogle Scholar
  41. Rivest RL (1997) S-expressions, Internet Engineering Task Force—Internet Draft. http://people.csail.mit.edu/rivest/Sexp.txt,1997
  42. Robilliard D, Marion-Poty V, Fonlupt C (2009) Genetic programming on graphics processing units. Genet Program Evol Mach 10(4):447–471CrossRefGoogle Scholar
  43. Russell SJ, Norvig P (2011) Artificial intelligence: a modern approach, 3rd edn. Pearson Education, BostonGoogle Scholar
  44. Smith SF (1980) A learning system based on genetic adaptive algorithms. PhD thesisGoogle Scholar
  45. Sutton RS, Barto AG (1998) Reinforcement learning: an introduction. MIT Press, CambridgeGoogle Scholar
  46. Tenne Y, Armfield S (2009) A framework for memetic optimization using variable global and local surrogate models. Soft Comput A Fusion Found Methodol Appl 13(8):781–793Google Scholar
  47. Thrun S (1996) Is learning the n-th thing any easier than learning the first? In: Advances in neural information processing systems. MIT Press, Cambridge, pp 640–646Google Scholar
  48. Urbanowicz RJ, Moore JH (2009) Learning classifier systems: a complete introduction, review, and roadmap. J Artif Evol Appl 2009(1):1–25CrossRefGoogle Scholar
  49. Wilson SW (1994) ZCS: a zeroth level classifier system. Evol Comput 2(1):1–18CrossRefGoogle Scholar
  50. Wilson SW (1995) Classifier fitness based on accuracy. Evol Comput 3(2):149–175CrossRefGoogle Scholar
  51. Wilson SW (1998) Generalization in the XCS classifier system. In: Procedings of the third annual genetic programming conference, pp 665–674Google Scholar
  52. Wilson SW (2002) Classifiers that approximate functions. Nat Comput 1:211–233MathSciNetzbMATHCrossRefGoogle Scholar

Copyright information

© Springer-Verlag 2012

Authors and Affiliations

  • Muhammad Iqbal
    • 1
    Email author
  • Will N. Browne
    • 1
  • Mengjie Zhang
    • 1
  1. 1.School of Engineering and Computer ScienceVictoria University of WellingtonWellingtonNew Zealand

Personalised recommendations