Abstract
Anticipatory Learning Classifier Systems (ALCS) are rule-based machine learning algorithms that can simultaneously develop a complete representation of their environment and a decision policy based on this representation to solve their learning tasks. This paper introduces BEACS (Behavioral Enhanced Anticipatory Classifier System) in order to handle non-deterministic partially observable environments and to allow users to better understand the environmental representations issued by the system. BEACS is an ALCS that enhances and merges Probability-Enhanced Predictions and Behavioral Sequences approaches used in ALCS to handle such environments. The Probability-Enhanced Predictions consist in enabling the anticipation of several states, while the Behavioral Sequences permits the construction of sequences of actions. The capabilities of BEACS have been studied on a thorough benchmark of 23 mazes and the results show that BEACS can handle different kinds of non-determinism in partially observable environments, while describing completely and more accurately such environments. BEACS thus provides explanatory insights about created decision policies and environmental representations.
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsReferences
Bagnall, A.J., Zatuchna, Z.V.: On the classification of maze problems. In: Bull, L., Kovacs, T. (eds.) Foundations of Learning Classifier Systems. Studies in Fuzziness and Soft Computing, pp. 305–316. Springer, Heidelberg (2005). https://doi.org/10.1007/11319122_12
Butz, A.M.V., Goldberg, B.D.E., Stolzmann, C.W.: The anticipatory classifier system and genetic generalization. Nat. Comput. 1, 427–467 (2002). https://doi.org/10.1023/A:1021330114221
Butz, M.V.: Biasing exploration in an anticipatory learning classifier system. In: Lanzi, P.L., Stolzmann, W., Wilson, S.W. (eds.) IWLCS 2001. LNCS (LNAI), vol. 2321, pp. 3–22. Springer, Heidelberg (2002). https://doi.org/10.1007/3-540-48104-4_1
Butz, M.V., Goldberg, D.E.: Generalized state values in an anticipatory learning classifier system. In: Butz, M.V., Sigaud, O., Gérard, P. (eds.) Anticipatory Behavior in Adaptive Learning Systems. LNCS (LNAI), vol. 2684, pp. 282–301. Springer, Heidelberg (2003). https://doi.org/10.1007/978-3-540-45002-3_16
Butz, M.V., Goldberg, D.E., Stolzmann, W.: Probability-enhanced predictions in the anticipatory classifier system. In: Luca Lanzi, P., Stolzmann, W., Wilson, S.W. (eds.) IWLCS 2000. LNCS (LNAI), vol. 1996, pp. 37–51. Springer, Heidelberg (2001). https://doi.org/10.1007/3-540-44640-0_4
Butz, M.V., Stolzmann, W.: An algorithmic description of ACS2. In: Lanzi, P.L., Stolzmann, W., Wilson, S.W. (eds.) IWLCS 2001. LNCS (LNAI), vol. 2321, pp. 211–229. Springer, Heidelberg (2002). https://doi.org/10.1007/3-540-48104-4_13
Fagerland, M.W., Sandvik, L.: Performance of five two-sample location tests for skewed distributions with unequal variances. Contemp. Clin. Trials 30, 490–496 (2009)
Gérard, P., Meyer, J.A., Sigaud, O.: Combining latent learning with dynamic programming in the modular anticipatory classifier system. Eur. J. Oper. Res. 160(3), 614–637 (2005)
Gerard, P., Stolzmann, W., Sigaud, O.: YACS: a new learning classifier system using anticipation. Soft Comput. 6, 216–228 (2002). https://doi.org/10.1007/s005000100117
Hasselt, H.: Double q-learning. In: Advances in neural information processing systems, pp. 2613–2621 (2010)
Hoffmann, J.: Anticipatory behavioral control. In: Butz, M.V., Sigaud, O., Gérard, P. (eds.) Anticipatory Behavior in Adaptive Learning Systems. LNCS (LNAI), vol. 2684, pp. 44–65. Springer, Heidelberg (2003). https://doi.org/10.1007/978-3-540-45002-3_4
Kozlowski, N., Unold, O.: Investigating exploration techniques for ACS in discretized real-valued environments. In: Proceedings of the 2020 Genetic and Evolutionary Computation Conference Companion, pp. 1765–1773 (2020)
Kozłowski, N., Unold, O.: Anticipatory classifier system with average reward criterion in discretized multi-step environments. Appl. Sci. 11(3), 1098 (2021)
Métivier, M., Lattaud, C.: Anticipatory classifier system using behavioral sequences in non-Markov environments. In: Lanzi, P.L., Stolzmann, W., Wilson, S.W. (eds.) IWLCS 2002. LNCS (LNAI), vol. 2661, pp. 143–162. Springer, Heidelberg (2003). https://doi.org/10.1007/978-3-540-40029-5_9
Orhand, R., Jeannin-Girardon, A., Parrend, P., Collet, P.: BACS: a thorough study of using behavioral sequences in ACS2. In: Bäck, T., et al. (eds.) PPSN 2020, Part I. LNCS, vol. 12269, pp. 524–538. Springer, Cham (2020). https://doi.org/10.1007/978-3-030-58112-1_36
Orhand, R., Jeannin-Girardon, A., Parrend, P., Collet, P.: PEPACS: integrating probability-enhanced predictions to ACS2. In: Proceedings of the 2020 Genetic and Evolutionary Computation Conference Companion, pp. 1774–1781 (2020)
Orhand, R., Jeannin-Girardon, A., Parrend, P., Collet, P.: Explainability and performance of anticipatory learning classifier systems in non-deterministic environments. In: Proceedings of the Genetic and Evolutionary Computation Conference Companion, pp. 163–164 (2021)
Stolzmann, W.: An introduction to anticipatory classifier systems. In: Lanzi, P.L., Stolzmann, W., Wilson, S.W. (eds.) IWLCS 1999. LNCS (LNAI), vol. 1813, pp. 175–194. Springer, Heidelberg (2000). https://doi.org/10.1007/3-540-45027-0_9
Stolzmann, W., Butz, M., Hoffmann, J., Goldberg, D.: First cognitive capabilities in the anticipatory classifier system, February 2000
Tan, J., Moore, J., Urbanowicz, R.: Rapid rule compaction strategies for global knowledge discovery in a supervised learning classifier system. In: Artificial Life Conference Proceedings, vol. 13, pp. 110–117 (2013)
Unold, O., Rogula, E., Kozłowski, N.: Introducing action planning to the anticipatory classifier system ACS2. In: Burduk, R., Kurzynski, M., Wozniak, M. (eds.) CORES 2019. AISC, vol. 977, pp. 264–275. Springer, Cham (2020). https://doi.org/10.1007/978-3-030-19738-4_27
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2022 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Orhand, R., Jeannin-Girardon, A., Parrend, P., Collet, P. (2022). Accurate and Interpretable Representations of Environments with Anticipatory Learning Classifier Systems. In: Medvet, E., Pappa, G., Xue, B. (eds) Genetic Programming. EuroGP 2022. Lecture Notes in Computer Science, vol 13223. Springer, Cham. https://doi.org/10.1007/978-3-031-02056-8_16
Download citation
DOI: https://doi.org/10.1007/978-3-031-02056-8_16
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-02055-1
Online ISBN: 978-3-031-02056-8
eBook Packages: Computer ScienceComputer Science (R0)