Advertisement

Decision trees and rule induction in simulated soccer agents

  • Ioan Alfred Letia
  • Marius Joldos
  • Calin Cenan
  • Diana Zaiu
  • Alina Andreica
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 1456)

Abstract

Low-level individual behavior and higher-level collaborative behavior are presented in a two-level layered architecture for simulated robotic soccer. Shooting to the goal and kicking in a passing situation are achieved by a neural network trained for various positions of the attacker and the goal-keeper or teammate, respectively. As collaboration with teammates passing is considered in various configurations of a group of four attackers and four defenders. Learning the decision of a player: (i) keep ball, (ii) pass ball to closest teammate, (iii) pass ball to medium distance teammate, (iv) pass ball to farthest teammate are studied. The algorithms used to learn this decision-making are OC1 and ITI, for decision tree, and CN2 and RIPPER for rule induction. These results are useful in constructing a decision-maker for a player.

Keywords

Robotic soccer Learning low-level behavior Learning higher level behavior Collaborative behavior with decision trees Rule induction 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Peter Clark and Robin Boswell. Rule induction with CN2: some recent improvements. In Yves Kodratoff, editor, Proceedings of the Fifth European Conference on Machine Learning, pages 151–163, 1991.Google Scholar
  2. 2.
    Peter Clark and Stan Matwin. Using qualitative models to guide inductive learning. In Paul Utgoff, editor, Proceedings of the 10th International Conference on Machine Learning, pages 49–56. Morgan Kaufmann, 1993.Google Scholar
  3. 3.
    William W. Cohen. Fast effective rule induction. In Proceedings of the Twelfth International Conference on Machine Learning, 1995.Google Scholar
  4. 4.
    Silvia Coradeschi and Lars Karlsson. A decision-mechanism for reactive and cooperating soccer-playing agents. Electronic Articles in Computer and Information Science 1, Linköping University, 1997.Google Scholar
  5. 5.
    Silvia Coradeschi and Lars Karlsson. A behavior-based approach to reactivity and coordination: a preliminary report. In Intelligent Agents, LNAI. Springer-Verlag: Heidelberg, Germany, 1998.Google Scholar
  6. 6.
    Silvia Coradeschi and Lars Karlsson. A role-based decision-mechanism for teams of reactive and coordinating agents. In RoboCup 97, LNAI. Springer-Verlag: Heidelberg, Germany, 1998.Google Scholar
  7. 7.
    H. Kitano, M. Tambe, P. Stone, S. Coradeschi, H. Matsubara, M. Veloso, I. Noda, E. Osawa, and M. Asada. The RoboCup synthetic agent challenge. In Proceedings of the Fifteenth International Joint Conference on Artificial Intelligence (IJCAI97), Yokohama, Japan, August 1997.Google Scholar
  8. 8.
    H. Matsubara, I. Noda, and K. Hiraki. Learning of cooperative actions in multiagent systems: A case study of pass play in soccer. In Adaptation, Coevolution and Learning in Multiagent Systems: AAAI Spring Symposium, pages 63–67, 1996.Google Scholar
  9. 9.
    S.K. Murthy, S. Kasif, and S. Salzberg. A system for induction of oblique decision trees. Journal of Artificial Intelligence Research, 2(1):1–32, 1994.Google Scholar
  10. 10.
    Itsuki Noda, Hitoshi Matsubara, and Kazuo Hiraki. Learning cooperative behavior in multi-agent environment: A case study of choice of play-plans in soccer. In Proceedings of PRICAI'96, pages 570–579, Cairas, Australia, August 1996.Google Scholar
  11. 11.
    Peter Stone and Manuela Veloso. A layered approach to learning client behaviors in the RoboCup soccer server. Applied Artificial Intelligence, 12, 1998.Google Scholar
  12. 12.
    Peter Stone and Manuela Veloso. Towards collaborative and adversarial learning: A case study in robotic soccer. International Journal of Human Computer Studies, 48, 1998.Google Scholar
  13. 13.
    Peter Stone and Manuela Veloso. Using decision tree confidence factors for multiagent control. In Hiroaki Kitano, editor, RoboCup-97: The First Robot World Cup Soccer Games and Conferences. Springer-Verlag: Heidelberg, Germany, 1998.Google Scholar
  14. 14.
    Milind Tambe. Towards flexible teamwork. Journal of Artificial Intelligence Research, 7:83–124, 1997.Google Scholar
  15. 15.
    Paul E. Utgoff, Neil C. Berkman, and Jeffery A. Clouse. Decision tree induction based on efficient tree restructuring. Machine Learning, pages 1–49, 1998.Google Scholar
  16. 16.
    Manuela Veloso, Peter Stone, and Kwun Han. The CMUnited-97 robotic soccer team: Perception and multiagent control. Technical report, Carnegie Mellon Univesity, October 1997.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1998

Authors and Affiliations

  • Ioan Alfred Letia
    • 1
  • Marius Joldos
    • 1
  • Calin Cenan
    • 1
  • Diana Zaiu
    • 1
  • Alina Andreica
    • 1
  1. 1.Department of Computer ScienceTechnical UniversityCluj-NapocaRomania

Personalised recommendations