COMBINING REINFORCEMENT LEARNING AND GENETIC ALGORITHMS TO LEARN BEHAVIOURS IN MOBILE ROBOTICS

R. Iglesias, M. Rodríguez, C. V. Regueiro, J. Correa, S. Barro

2006

Abstract

Reinforcement learning is an extremely useful paradigm which is able to solve problems in those domains where it is difficult to get a set of examples of how the system should work. Nevertheless, there are important problems associated with this paradigm which make the learning process more unstable and its convergence slower. In our case, to overcome one of the main problems (exploration versus exploitation trade off), we propose a combination of reinforcement learning with genetic algorithms, where both paradigms influence each other in such a way that the drawbacks of each paradigm are balanced with the benefits of the other. The application of our proposal to solve a problem in mobile robotics shows its usefulness and high performance, as it is able to find a stable solution in a short period of time. The usefulness of our approach is highlighted through the application of the system learnt through our proposal to control the real robot.

References

  1. Cichosz, P. (1997). Reinforcement Learning by Truncating Temporal Differences. PhD thesis, Dpt. of Electronics and Information Technology, Warsaw University of Technology.
  2. Cobb, H. G. and Grefenstette, J. J. (1993). Genetic algorithms for tracking changing environments. In Proc. Fifth International Conference on Genetic Algorithms.
  3. David E. Moriarty, Alan C. Schultz, J. J. G. (1999). Evolutionary algorithms for reinforcement learning. Journal of Artificial Intelligence Research, 11:241-276.
  4. Davidor, Y. (1991). Genetic algorithms and robotics. A heuristic strategy for optimization. World Scientific.
  5. Grefenstette, J. J. (1992). Parallel Problem Solving from Nature, volume 2, chapter Genetic algorithms for changing environments.
  6. R. Iglesias, C. V. Regueiro, J. C. and Barro, S. (1998a). Improving wall following behaviour in a mobile robot using reinforcement learning. In ICSC International Symposium on Engineering of Intelligent Systems, EIS'98.
  7. R. Iglesias, C. V. Regueiro, J. C. and Barro, S. (1998b). Supervised reinforcement learning: Application to a wall following behaviour in mobile robotics. In International Conference on Ingustrial & Engineering Applications of Artificial Intelligence & Expert Systems.
  8. R. Iglesias, M. F. D. and Barro, S. Learning of perceptual states in the design of an adaptive wall-following behaviour. In European Symposium on Artificial Neural Networks, ESANN 2000.
Download


Paper Citation


in Harvard Style

Iglesias R., Rodríguez M., V. Regueiro C., Correa J. and Barro S. (2006). COMBINING REINFORCEMENT LEARNING AND GENETIC ALGORITHMS TO LEARN BEHAVIOURS IN MOBILE ROBOTICS . In Proceedings of the Third International Conference on Informatics in Control, Automation and Robotics - Volume 2: ICINCO, ISBN 978-972-8865-60-3, pages 188-195. DOI: 10.5220/0001209501880195


in Bibtex Style

@conference{icinco06,
author={R. Iglesias and M. Rodríguez and C. V. Regueiro and J. Correa and S. Barro},
title={COMBINING REINFORCEMENT LEARNING AND GENETIC ALGORITHMS TO LEARN BEHAVIOURS IN MOBILE ROBOTICS},
booktitle={Proceedings of the Third International Conference on Informatics in Control, Automation and Robotics - Volume 2: ICINCO,},
year={2006},
pages={188-195},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0001209501880195},
isbn={978-972-8865-60-3},
}


in EndNote Style

TY - CONF
JO - Proceedings of the Third International Conference on Informatics in Control, Automation and Robotics - Volume 2: ICINCO,
TI - COMBINING REINFORCEMENT LEARNING AND GENETIC ALGORITHMS TO LEARN BEHAVIOURS IN MOBILE ROBOTICS
SN - 978-972-8865-60-3
AU - Iglesias R.
AU - Rodríguez M.
AU - V. Regueiro C.
AU - Correa J.
AU - Barro S.
PY - 2006
SP - 188
EP - 195
DO - 10.5220/0001209501880195