We consider the problem of computing the expected accumulated reward and the average gain per transition in a subclass of Markov chains with countable state spaces where all states are assigned a non-negative reward. We state several abstract conditions that guarantee computability of the above properties up to an arbitrarily small (but non-zero) given error. Finally, we show that our results can be applied to probabilistic lossy channel systems, a well-known model of processes communicating through faulty channels. © Springer-Verlag Berlin Heidelberg 2005.
CITATION STYLE
Brázdil, T., & Kučera, A. (2005). Computing the expected accumulated reward and gain for a subclass of infinite Markov chains. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 3821 LNCS, pp. 372–383). Springer Verlag. https://doi.org/10.1007/11590156_30
Mendeley helps you to discover research relevant for your work.