Learning on real robots from experience and simple user feedback

• Autores: Pablo Quintía Vidal, Roberto Iglesias Rodríguez , Miguel Ángel Rodríguez González , Carlos V. Regueiro
• Localización: JoPha: Journal of Physical Agents, ISSN-e 1888-0258, Vol. 7, Nº. 1, 2013 (Ejemplar dedicado a: Special issue on advances on physical agents), pág. 8
• Idioma: inglés
• DOI: 10.14198/jopha.2013.7.1.08
• Enlaces
  • Texto completo
• Resumen

  • In this article we describe a novel algorithm that allows fast and continuous learning on a physical robot working in a real environment. The learning process is never stopped and new knowledge gained from robot-environment interactions can be incorporated into the controller at any time. Our algorithm lets a human observer control the reward given to the robot, hence avoiding the burden of defining a reward function.

    Despite the highly-non-deterministic reinforcement, through the experimental results described in this paper, we will see how the learning processes are never stopped and are able to achieve fast robot adaptation to the diversity of different situations the robot encounters while it is moving in several environments

• Referencias bibliográficas
  • [1] R. S. Sutton. Reinforcement learning: An introduction. MIT Press, 1998.
  • [2] M.A. Bozarth. Pleasure: The politics and the reality, Springer Netherlands, pp. 5–14, 1994.
  • [3] E.L. Thorndike. Animal Intelligence, Hafner, Darien, CT, 1911.
  • [4] A. L. Thomaz and C. Breazeal. “Teachable robots: Understanding human teaching behaviour to build more effective robot learners.” Artificial...
  • [5] A. L. Thomaz, G. Hoffman, and C. Breazeal. “Reinforcement learning with human teachers: Understanding how people want to teach robots.”...
  • [6] L.P. Kaelbling, M.L. Littman, and A.W. Moore. “Reinforcement learning: A survey.” Journal of Artificial Intelligence Research, 4:237–285,...
  • [7] G. A. Carpenter, S. Grossberg, and D. B. Rosen. “Fuzzy ART: Fast stable learning and categorization of analog patterns by an adaptive...
  • [8] Pablo Quintia, Roberto Iglesias, Carlos V. Regueiro, and Miguel A. Rodriguez. “Simultaneous learning of perception and action in mobile...
  • [9] W. Bradley Knox and Peter Stone. “Interactively shaping agents via human reinforcement: The TAMER framework.” In Fifth International Conference...
  • [10] W. Bradley Knox and Peter Stone. “Reinforcement learning with human feedback in mountain car.” In AAAI Spring 2011 Symposium on Bridging...
  • [11] Andrew Y. Ng and Stuart Russell. “Algorithms for inverse reinforcement learning.” In Proceedings of the Seventeenth International Conference...
  • [12] Maass W., Natschlaeger T., and Markram H. “Real-time computing without stable states: A new framework for neural computation based on...
  • [13] Brenna D. Argall, Sonia Chernova, Manuela Veloso, and Brett Browning. “A survey of robot learning from demonstration.” Robotics and Autonomous...
  • [14] Andrea Lockerd Thomaz, Guy Hoffman, and Cynthia Breazeal. “Real-time interactive reinforcement learning for robots.” In AAAI Workshop...
  • [15] T. Kollar and N. Roy. “Using reinforcement learning to improve exploration trajectories for error minimization.” In IEEE International...