A learning algorithm for discrete-time stochastic control

Borkar, V. S. (2000) A learning algorithm for discrete-time stochastic control Probability in the Engineering and Informational Sciences, 14 (2). pp. 243-258. ISSN 0269-9648

Full text not available from this repository.

Official URL: http://portal.acm.org/citation.cfm?id=984613.98462...

Related URL: http://dx.doi.org/10.1017/S0269964800142081


A simulation-based algorithm for learning good policies for a discrete-time stochastic control process with unknown transition law is analyzed when the state and action spaces are compact subsets of Euclidean spaces. This extends the Q-learning scheme of discrete state/action problems along the lines of Baker [4]. Almost sure convergence is proved under suitable conditions.

Item Type:Article
Source:Copyright of this article belongs to Cambridge University Press.
ID Code:5317
Deposited On:18 Oct 2010 08:38
Last Modified:20 May 2011 09:07

Repository Staff Only: item control page