Abstract
This paper describes a novel methodology for behavior learning of an agent, called “Coaching”. This is an interactive learning method which allows a human trainer to give a subjective evaluation to the robotic agent in real time. and the agent can update the reward function dynamically based on the evaluation. We demonstrated that the agent is capable of learning the desired behavior by being given simple and subjective instruction such as “good and bad”, The proposed approach is also effective when it is difficult to determine a suitable reward function for the learning situation in advance.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Sutton, R.S., Barto, A.G.: Reinforcement Learning: An Introduction, A Bradford Book. MIT Press, Cambridge (1998)
Inamura, T., Tanie, H., Nakamura, Y.: From Stochastic Motion Generation and Recognition to Geometric Symbol Development and Manipulation. In: Proc. of IEEE-RAS Int’l. Conf. on Humanoid Robots (2003)
Young, J.E., Igarashi, T., Sharlin., E.: Puppet Master: Designing Reactive Character Behavior by Demonstration. In: Proc. SCA 2008, pp. 183–191 (2008)
Nakatani, M., Suzuki, K., Hashimoto, S.: Subjective-Evaluation Oriented Teaching Scheme for a Biped Humanoid Robot. In: Proc. of IEEE-RAS Intl. Conf. on Humanoid Robots (2003)
Riley, M., Ude, A., Atkeson, C., Chang, G.: Coaching: An Approach to Efficiently and Intuitively Create Humanoid Robot Behaviors. In: Proc. of IEEE-RAS Intl. Conf. on Humanoid Robots, pp. 567–574 (2006)
Doya, K.: Reinforcement learning in continuous time and space. Neural Computation 12, 219–245 (2000)
Kamatani, H., Kitayama, K., Fujimura, A., Abe, K.: Reinforcement Learning in Continuous State Space. In: SICE Tohoku Chapter Workshop, pp. 229–11 (2006) (in Japanese)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2010 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Hirokawa, M., Suzuki, K. (2010). Coaching to Enhance the Online Behavior Learning of a Robotic Agent. In: Setchi, R., Jordanov, I., Howlett, R.J., Jain, L.C. (eds) Knowledge-Based and Intelligent Information and Engineering Systems. KES 2010. Lecture Notes in Computer Science(), vol 6276. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-15387-7_19
Download citation
DOI: https://doi.org/10.1007/978-3-642-15387-7_19
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-642-15386-0
Online ISBN: 978-3-642-15387-7
eBook Packages: Computer ScienceComputer Science (R0)