SELF CONSTRUCTING NEURAL NETWORK ROBOT CONTROLLER BASED ON ON-LINE TASK PERFORMANCE FEEDBACK

Andreas Huemer, Mario Gongora, David Elizondo

Abstract

A novel methodology to create a powerful controller for robots that minimises the design effort is presented. We show that using the feedback from the robot itself, the system can learn from experience. A method is presented where the interpretation of the sensory feedback is integrated in the creation of the controller, which is achieved by growing a spiking neural network system. The feedback is extracted from a performance measuring function provided at the task definition stage, which takes into consideration the robot actions without the need for external or manual analysis.

References

  1. Daucé, E. and Henry, F. (2006). Hebbian learning in large recurrent neural networks. Technical report, Movement and Perception Lab, Marseille.
  2. Elizondo, D., Fiesler, E., and Korczak, J. (1995). Nonontogenetic sparse neural networks. In International Conference on Neural Networks 1995, IEEE, volume 26, pages 290-295.
  3. Florian, R. V. (2005). A reinforcement learning algorithm for spiking neural networks. In Proceedings of the Seventh International Symposium on Symbolic and Numeric Algorithms for Scientific Computing, pages 299-306.
  4. Gers, F. A., Schraudolph, N. N., and Schmidhuber, J. (2002). Learning precise timing with LSTM recurrent networks. Journal of Machine Learning Research, 3:115-143.
  5. Gómez, G., Lungarella, M., Hotz, P. E., Matsushita, K., and Pfeifer, R. (2004). Simulating development in a real robot: On the concurrent increase of sensory, motor, and neural complexity. In Proceedings of the Fourth International Workshop on Epigenetic Robotics, pages 119-122.
  6. Hebb, D. O. (1949). The Organization of Behaviour: A Neuropsychological Approach. John Wiley & Sons, New York.
  7. Izhikevich, E. M. (2006). Polychronization: Computation with spikes. Neural Computation, 18:245-282.
  8. Izhikevich, E. M. (2007). Solving the distal reward problem through linkage of STDP and dopamine signaling. Cerebral Cortex, 10:1093-1102.
  9. Katic, D. (2006). Leaky-Integrate-and-Fire und Spike Re-
  10. nische Informatik, Universitï¿ 12 t Karlsruhe.
  11. sponse Modell. Technical report, Institut fï¿ 12 r Tech-
  12. Liu, J. and Buller, A. (2005). Self-development of motor abilities resulting from the growth of a neural network reinforced by pleasure and tension. In Proceedings of the 4th International Conference on Development and Learning 2005, pages 121-125.
  13. Liu, J., Buller, A., and Joachimczak, M. (2006). Selfmotivated learning agent: Skill-development in a growing network mediated by pleasure and tensions. Transactions of the Institute of Systems, Control and Information Engineers, 19(5):169-176.
  14. van Leeuwen, M. (2004). Spike timing dependent structural plasticity in a single model neuron. Master's thesis, Intelligent Systems Group, Institute for Information and Computing Sciences, Utrecht University.
  15. Vreeken, J. (2003). Spiking neural networks, an introduction. Technical report, Intelligent Systems Group, Institute for Information and Computing Sciences, Utrecht University.
Download


Paper Citation


in Harvard Style

Huemer A., Gongora M. and Elizondo D. (2008). SELF CONSTRUCTING NEURAL NETWORK ROBOT CONTROLLER BASED ON ON-LINE TASK PERFORMANCE FEEDBACK . In Proceedings of the Fifth International Conference on Informatics in Control, Automation and Robotics - Volume 1: ICINCO, ISBN 978-989-8111-30-2, pages 326-333. DOI: 10.5220/0001495503260333


in Bibtex Style

@conference{icinco08,
author={Andreas Huemer and Mario Gongora and David Elizondo},
title={SELF CONSTRUCTING NEURAL NETWORK ROBOT CONTROLLER BASED ON ON-LINE TASK PERFORMANCE FEEDBACK},
booktitle={Proceedings of the Fifth International Conference on Informatics in Control, Automation and Robotics - Volume 1: ICINCO,},
year={2008},
pages={326-333},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0001495503260333},
isbn={978-989-8111-30-2},
}


in EndNote Style

TY - CONF
JO - Proceedings of the Fifth International Conference on Informatics in Control, Automation and Robotics - Volume 1: ICINCO,
TI - SELF CONSTRUCTING NEURAL NETWORK ROBOT CONTROLLER BASED ON ON-LINE TASK PERFORMANCE FEEDBACK
SN - 978-989-8111-30-2
AU - Huemer A.
AU - Gongora M.
AU - Elizondo D.
PY - 2008
SP - 326
EP - 333
DO - 10.5220/0001495503260333