Abstract
Recently, an efficient reinforcement learning method has been proposed, in which the problem of approximating the value function is naturally decomposed into a number of sub-problems, each of which can be solved at small computational cost. While this method certainly reduces the magnitude of temporal difference error, the value function may be overfitted to sampled data. To overcome this difficulty, we introduce a robust approximation to this context. Computer experiments show that the value function learning by our method is much more robust than those by the previous methods.
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
Bertsekas, D., Castañon, D.: Adaptive aggregation methods for infinite horizon dynamic programming. IEEE Transactions on Automatic Control 34, 589–598 (1989)
Bradtke, S.J., Barto, A.G.: Linear least-squares algorithms for temporal difference learning. Machine Learning 22(2), 33–57 (1996)
Dasgupta, M., Mishra, S., Hill, N.: Least absolute deviation estimation of linear economic models: A literature review. Munich Personal PePEc Archive 1781 (2004)
Keller, P.W., Mannor, S., Precup, D.: Automatic basis function construction for approximate dynamic programming and reinforcement learning. In: Proceedings of the Twenty-third International Conference on Machine Learning (2006)
Mori, T., Ishii, S.: An additive reinforcement learning. In: The Ninteenth International Conference on Artificial Neural Networks (2009)
Sutton, R.S., Barto, A.G.: Reinforcement Learning: An Introduction. MIT Press, Cambridge (1998)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2009 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Mori, T., Ishii, S. (2009). Robust Approximation in Decomposed Reinforcement Learning. In: Leung, C.S., Lee, M., Chan, J.H. (eds) Neural Information Processing. ICONIP 2009. Lecture Notes in Computer Science, vol 5863. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-10677-4_67
Download citation
DOI: https://doi.org/10.1007/978-3-642-10677-4_67
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-642-10676-7
Online ISBN: 978-3-642-10677-4
eBook Packages: Computer ScienceComputer Science (R0)