Abstract
The anticipatory classifier system (ACS)combines the learning classifier system frameworkwith the cognitive learning theory ofanticipatory behavioral control. The result is an evolutionary system thatbuilds a complete and generalized predictiveenvironmental model. Reinforcement learningtechniques are applied to form a behavioralpolicy represented in the model. After providingsome background as well as outlining the objectives of the system, we explainin detail all involved current processes. Furthermore, we analyze thedeficiency of over-specialization in the anticipatory learning process (ALP),the main learning mechanism in the ACS. Consequently, we introduce a geneticalgorithm (GA) to the ACS that is meant for generalization of over-specializedclassifiers. We show that it is possible to form a symbiosis between a directedspecialization and a genetic generalization mechanism achieving a learningmechanism that evolves a complete, accurate, and compact description of theperceived environment. Results in three different environmental settingsconfirm the usefulness of the genetic algorithm in the ACS. Finally, we discuss future research directions.
Similar content being viewed by others
References
Booker LB, Goldberg DE and Holland JH (1989) Classifier systems and genetic algorithms. Artificial Intelligence 40: 235–282
Butz MV, Goldberg DE and Stolzmann W (2000a) Introducing a genetic generalization pressure to the anticipatory classifier system: Part 1-theoretical approach. In: Whitely D, Goldberg DE, Cantu-Paz E, Spector L, Parmee I and Beyer H-G (eds) Proceedings of the Genetic and Evolutionary Computation Conference (GECCO-2000), pp. 34–41. Morgan Kaufmann, San Francisco, CA
Butz MV, Goldberg DE and Stolzmann W (2000b) Introducing a genetic generalization pressure to the anticipatory classifier system: Part 2-performance analysis. In: Whitely D, Goldberg DE, Cantu-Paz E, Spector L, Parmee I and Beyer H-G (eds) Proceedings of the Genetic and Evolutionary Computation Conference (GECCO-2000), pp. 42–49. Morgan Kaufmann, San Francisco, CA
Butz MV, Goldberg DE and Stolzmann W (2000c) Investigating genetic generalization in the anticipatory classifier system. In: Schoenauer M, Deb K, Rudolph G, Yao X, Lutton E, Merely JJ and Schwefel H-P (eds) Parallel Problem Solving from Nature-PPSN VI, LNCS 1917, pp. 735–744. Springer-Verlag, Berlin
Butz MV, Goldberg DE and Stolzmann W (2001) Probability-enhanced predictions in the anticipatory classifier system. In: Lanzi PL, Stolzmann Wand Wilson SW (eds) Advances in Learning Classifier Systems-Third International Workshop, IWLCS-200, LNAI 1996, pp. 38–52. Springer-Verlag, Berlin. in press
Cliff D and Ross S (1994) Adding temporary memory to ZCS. Adaptive Behavior 3(2): 101–150
Dorigo M and Colombetti M (1998) Robot Shaping, an Experiment in Behavior Engineering, Intelligent Robotics and Autonomous Agents. MIT press, Cambridge, MA
Drescher GL (1991) Made-upMinds, a Constructivist Approach to Artificial Intelligence. MIT Press, Cambridge, MA
Goldberg D and Richardson J (1987) Genetic algorithms with sharing for mulitmodal function optimization. In: Grefenstette JJ (ed) Proceedings of the Second International Conference on Genetic Algorithms, pp. 41–49. Lawrence Erlbaum Associates, Hillsdale, NJ
Goldberg DE (2002) The Design of Innovation: Lesssons from and for Competent Genetic Algorithms, Genetic Algorithms and Evolutionary Computation. Kluwer Academic Publishers, Boston, MA
Goldberg DE and Deb K (1991) A comparative analysis of selection schemes used in genetic algorithms. Foundations of Genetic Algorithms 1: 69–93 (Also TCGA Report 90007)
Grefenstette JJ (1988) Credit assignment in rule discovery systems based on genetic algorithms. Machine Learning 3: 225–245.
Hoffmann J (1993) Vorhersage und Erkenntnis [Anticipation and Cognition]. Hogrefe, Goettingen, Germany Holland JH (1985) Properties of the bucket brigade algorithm. In: John J. Grefenstette (ed) Proceedings of an International Conference on Genetic Algorithms and their Applications, pp. 1–7. Carnegie-Mellon University, Pittsburgh, PA
Holland JH (1990) Concerning the emergence of tag-mediated lookahead in classifier systems. In: Forrest S (ed) Emergent Computation, pp. 188–201. North-Holland
Holland JH (1992) Adaptation in Natural and Artificial Systems: An Introductory Analysis with Applications to Biology, Control, and Artificial Intelligence (Complex A).MIT press
Holland JH and Reitman JS (1978) Cognitive systems based on adaptive algorithms. In: Waterman DA and Hayes-Roth F (eds) Pattern Directed Inference Systems, pp. 313–329. Academic Press, New York
James W (1981) (orig. 1890) The Principles of Psychology (Vol. 2). Harvard University Press, Cambridge, MA
Kovacs T (1996) Evolving Optimal Populations with XCS Classifier Systems. Master's thesis, School of Computer Science, University of Birmingham, Birmingham, UK
Kovacs T (1997) XCS classifier system reliably evolves accurate, complete, and minimal representations for boolean functions. In: Roy Chawdhry and Pant (eds) Soft Computing in Engineering Design and Manufacturing, pp. 59–68. Springer-Verlag, London
Kuvayev L and Sutton RS (1996) Model-based reinforcement learning with an approximate, learned model. In: Proceedings of the Ninth Yale Workshop on Adaptive and Learning Systems, pp. 101–105. New Have, CT
LaBerge D (1995) Attentional Processing, the Brain's Art of Mindfulness. Harvard University Press, Cambridge, MA
Lanzi PL (1999) An analysis of generalization in the XCS classifier system. Evolutionary Computation 7(2): 125–149
Lanzi PL (2000a) Adaptive agents with reinforcemen learning and internal memory. In: Meyer JA, Berthoz A, Floreano D, Roitblat H and Wilson SW (eds) From Animals to Animats 6: Proceedings of the Sixth International Conference on Simulation of Adaptive Behavior, pp. 333–342. MIT Press, Cambridge, MA
Lanzi PL (2000b) Learning Classifier Systems from a Reinforcement Learning Perspective. Technical Report 00–03, Dipartimento di Elettronica e Informazione, Politecnico di Milano
Lanzi PL and Wilson SW (2000) Toward optimal classifier system performance in non-markov environments. Evoultionary Computation 8(4): 393–418
Riolo RL (1991) Lookahead planning and latent learning in a classifier system. In: Meyer JA and Wilson SW (eds) From Animals to Animats: Proceedings of the First International Conference on Simulation of Adaptive Behavior, pp. 316–326. MIT Press, Cambridge, MA
Rosen R (1985) Anticipatory Systems. Pergamon, Oxford, UK
Rosen R (1991) Life Itself (Complexity in Ecological Systems Series). Columbia University Press, New York, USA
Stolzmann W (1997) Antizipative Classifier Systeme [Anticipatory Classifier Systems]. Shaker Verlag, Aachen, Osnabrueck, Germany
Stolzmann W (1998) Anticipatory classifier systems. In: Genetic Programming '98, pp. 658–664. University of Wisconsin, Morgan Kaufmann, Madison, Wisconsin
Stolzmann W (2000) An introduction to anticipatory classifier systems. In: Lanzi PL, Stolzmann W and Wilson SW (eds) Learning Classifier Systems: From Foundations to Applications, LNAI 1813, pp. 175–194. Springer-Verlag, Berlin
Stolzmann W and Butz MV (2000) Latent learning and action-planning in robots with anticipatory classifier systems. In: Lanzi PL, Stolzmann W and Wilson SW (eds) Learning Classifier Systems: From Foundations to Applications, LNAI 1813, pp. 301–317. Springer-Verlag, Berlin
Stolzmann W, Butz MV, Hoffmann J and Goldberg DE (2000) First cognitive capabilities in the anticipatory classifier system. In: Meyer J-A, Berthoz A, Floreano D, Roitblat H and Wilson SW (eds) From Animals to Animats 6: Proceedings of the Sixth International Conference on Simulation of Adaptive Behavior, pp. 287–296. MIT press, Cambridge, MA
Sutton RS (1991a) DYNA, an integrated architecture for learning, planning, and reacting. In: Working Notes of the 1991 AAAI Spring Symposium on Integrated Intelligent Architectures, pp. 151–155
Sutton RS (1991b) Reinforcement learning architectures for animats. In: Meyer J-A and Wilson SW (eds) From Animals to Animats: Proceedings of the First International Conference on Simulation of Adaptive Behavior, pp. 288–296. MIT Press, Cambridge, MA
Sutton RS and Barto AG (1998) Reinforcement Learning: An Introduction. MIT Press, Cambridge, MA
Thislethwaite D (1951) A critical review of latent learning and related experiments. Psychological Bulletin 48(2): 97–129
Tolman EC (1932) Purposive Behavior in Animals and Men. Appleton, New York Tomlinson A and Bull L (2001) Symbiogenesis in learning classifier systems: Corporations, niches and persistence. Artificial Life. in press
Watkins CJCH and Dayan P (1992) Q-learning. Machine Learning 8(3): 272–292
Whitehead SD and Ballard DH (1991) Learning to perceive and act. Machine Learning 7(1): 45–83
Widrow B and Hoff M (1960) Adaptive switching circuits. Western Electronic Show and Convention 4: 96–104
Wilson SW (1994) ZCS: A zeroth level classifier system. Evolutionary Computation 2(1): 1–18
Wilson SW (1995) Classifier fitness based on accuracy. Evolutionary Computation 3(2): 149–175
Wilson SW (1998) Generalization in the XCS classifier system. In: Koza JR, Banzhaf W, Chellapilla K, Deb K, DorigoM, Fogel DB, GarzonMH, Goldberg DE, Iba H and Riolo R (eds) Genetic Programming 1998: Proceedings of the Third Annual Conference, pp. 665–674. Morgan Kaufmann, San Francisco
Wilson SW (2000) Get real! XCS with continuous-valued inputs. In: Lanzi PL, Stolzmann W and Wilson SW (eds) Learning Classifier Systems: From Foundations to Applications, LNAI 1813, pp. 209–219. Springer-Verlag, Berlin Heidelberg
Wilson SW and Goldberg DE (1989) A critical review of classifier systems. In: Schaffer JD (ed) Proceedings of the Third International Conference on Genetic Algorithms, pp. 244–255. Morgan Kaufmann, San Mateo, CA
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Butz, A.M.V., Goldberg, B.D.E. & Stolzmann, C.W. The anticipatory classifier system and genetic generalization. Natural Computing 1, 427–467 (2002). https://doi.org/10.1023/A:1021330114221
Issue Date:
DOI: https://doi.org/10.1023/A:1021330114221