Advertisement

Reinforcement Learning with Raw Image Pixels as Input State

  • Damien Ernst
  • Raphaël Marée
  • Louis Wehenkel
Part of the Lecture Notes in Computer Science book series (LNCS, volume 4153)

Abstract

We report in this paper some positive simulation results obtained when image pixels are directly used as input state of a reinforcement learning algorithm. The reinforcement learning algorithm chosen to carry out the simulation is a batch-mode algorithm known as fitted Q iteration.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Ernst, D., Geurts, P., Wehenkel, L.: Tree-based batch mode reinforcement learning. Journal of Machine Learning Research 6, 503–556 (2005)MathSciNetGoogle Scholar
  2. 2.
    Geurts, P., Ernst, D., Wehenkel, L.: Extremely randomized trees. Machine Learning 36(1), 3–42 (2006)CrossRefGoogle Scholar
  3. 3.
    Jodogne, S., Piater, S.: Interactive learning of mappings from visual percepts to actions. In: De Raedt, L., Wrobel, S. (eds.) Proceedings of the 22nd International Conference on Machine Learning, pp. 393–400 (August 2005)Google Scholar
  4. 4.
    Lagoudakis, M., Parr, R.: Reinforcement learning as classification: leveraging modern classifiers. In: Faucett, T., Mishra, N. (eds.) Proceedings of 20th International Conference on Machine Learning, pp. 424–431 (2003)Google Scholar
  5. 5.
    Marée, R., Geurts, P., Piater, J., Wehenkel, L.: Random subwindows for robust image classification. In: Schmid, C., Soatto, S., Tomasi, C. (eds.) Proceedings of the IEEE International Conference on Computer Vision and Pattern Recognition, vol. 1, pp. 34–40. IEEE, Los Alamitos (June 2005)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2006

Authors and Affiliations

  • Damien Ernst
    • 1
  • Raphaël Marée
    • 1
  • Louis Wehenkel
    • 1
  1. 1.Department of Electrical Engineering and Computer ScienceInstitut Montefiore – University of LiègeLiègeBelgium

Personalised recommendations