Abstract
The aim of behavioural cloning is to synthesize artificial controllers that are robust and comprehensible to human understanding. To attain the two objectives we propose the use of the Incremental Correction model that is based on a closed-loop control strategy to model the reactive aspects of human control skills. We have investigated the use of three different representations to encode the artificial controllers: univariate decision trees as induced by C4.5; multivariate decision and regression trees as induced by cart and; clausal theories induced by an Inductive Logic Programming (ILP) system.
We obtained an increase in robustness and a lower complexity of the controllers when compared with results using other models. The controllers synthesized by cart revealed to be the most robust. The ILP system produced the simpler encodings.
Chapter PDF
Similar content being viewed by others
References
M. Bain and C. Sammut. A framework for behavioural cloning. In Machine Intelligence 15. Oxford University Press, Oxford, U.K., 1999. (to appear).
R. Camacho. Inducing models of human control skills. In Proceedings of the European Conference on Machine Learning — ECML-98, Germany, April 1998.
R. Camacho. Inducing Models of Human Control Skills using Machine Learning Algorithms. PhD thesis, Universidade do Porto, July 2000.
A. A. Covrigaru and R. K. Lindsay. Deterministic autonomous systems. AI Magazine, 12(3):110–117, fall 1991.
J. C. Hamm. The use of pilot models in dynamic performance and rotor load prediction studies. In Proceedings of the Eighteenth European Rotorcraft Forum, pages 15–18, Avignon, France, September 1992. Association Aeronautique et Astronautique de France.
H. G. John, R. Kohavi, and K. Pfleger. Irrelevant features and the subset selection proble. In W. W. Cohen and H. Hirsh, editors, Machine Learning: Proceedings of the Eleventh International Conference, pages 121–129, San Francisco, California, June 1994. Morgan Kaufmann.
D. Michie, M. Bain, and J. Hayes-Michie. Cognitive models from subcognitive skills. In M. G. J. McGhee and P. Mowforth, editors, Knowledge-Based Systems for Industrial Control, pages 71–99. Peter Peregrinus for IEE, London, UK, 1990.
D. Michie and R. Camacho. Building symbolic representations of intuitive real-time skills from performance data. In D. M. eds. K. Furukawa and S. Muggleton, editors, Machine Intelligence 13, pages 385–418. Oxford University Press, Oxford, United Kingdom, 1994.
D. Michie and R. A. Chambers. Boxes: an experiment in adaptive control. In Machine Intelligence 2, pages 137–152. Oliver and Boyd, Edinburgh, 1968. eds. Dale, E. and Michie, Donald.
J. Randlov and P. Alstrom. Learning to drive a bicycle using reinforcement learning and shaping. In Proceedings of the International Conference on Machine Learning — ICML-98, pages 463–471, Madison, Wisconsin USA, July 1998.
M. Ryan and M. Reid. Learning to fly: An application of hierarchical reinforcement learning. In P. e. Langley, editor, Proceedings of the Seventeenth International Machine Learning Conference, ICML-2000, pages 807–814, San Francisco, CA., 2000. Morgan Kaufmann Publishers.
Y. Sakawa and Y. Shinido. Optimal control of container cranes. Automatica, 18:257–266, 1982.
C. Sammut. Experimental results from an evaluation of algorithms that learn to control dynamic systems. In Proceedings of the Fifth International Workshop of Machine Learning 88, pages 437–443, Ann Arbor, Univ of Michigan, June 1988. editor John Laird.
C. Sammut and J. Cribb. Is learning rate a good performance criterion for learning? In Proceedings of the Seventh International Workshop of Machine Learning 90, pages 170–178, Texas, June 1990.
C. Sammut, S. Hurst, D. Kedzier, and D. Michie. Learning to fly. In Proceedings of the Ninth International Workshop of Machine Learning 92, pages 385–393, Aberdeen, U.K., 1992.
D. Stirling. CHURPs: Compressed Heuristic Reaction Planners. PhD thesis, University of Sydney, 1995.
D. Šuc and I. Bratko. Skill reconstruction as induction of lq controllers with sub-goals. In Proceedings of the Fifteenth International Joint Conference on Artificial Intelligence-JCAI-97, volume 2, pages 914–920, Nagoya, Japan, 1997.
T. Urbančič and I. Bratko. Reconstructing human skill with machine learning. In The Eleventh European Conference on Artificial Intelligence, pages 498–502, Amsterdam, Netherlands, 1994. ed. A. Chon.
T. Urbančič and I. Bratko. Controlling container cranes: A case-study in reconstruction of human skill. In The Second International Workshop on Artificial Intelligence Techniques— AIT95, pages 113–126, Brno, Czech Republic, 1995.
D. Whitley. Genetic algorithms and neural networks. In Genetic Algorithms in Engineering and Computer Science, chapter 11. John Wiley & Sons Ltd, 1995. Eds. J. Periaux and G. Winter.
B. Widrow, D. E. Rumelhart, and M. A. Lehr. Neural networks: Applications in industry, business and science. Communications of the ACM, 37(3):93–105, 1994.
S. Yasunobu and T. Hasegawa. Evaluation of an automatic container crane operation system based on predictive fuzzy control. Control-Theory and Advanced Technology, 2:419–432, 1986.
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2003 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Camacho, R., Brazdil, P. (2003). Improving the Robustness and Encoding Complexity of Behavioural Clones. In: De Raedt, L., Flach, P. (eds) Machine Learning: ECML 2001. ECML 2001. Lecture Notes in Computer Science(), vol 2167. Springer, Berlin, Heidelberg. https://doi.org/10.1007/3-540-44795-4_4
Download citation
DOI: https://doi.org/10.1007/3-540-44795-4_4
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-42536-6
Online ISBN: 978-3-540-44795-5
eBook Packages: Springer Book Archive