Motivated by the computational difficulty of analyzing very large Markov chains, we define a notion of clusters in (not necessarily reversible) Markov chains, and explore the possibility of analyzing a cluster "in vitro," without regard to the remainder of the chain. We estimate the stationary probabilities of the states in the cluster using only transition information for these states, and bound the error of the estimate in terms of parameters measuring the quality of the cluster. Finally, we relate our results to searching in a hyperlinked environment, and provide supporting experimental results. © Springer-Verlag Berlin Heidelberg 2006.
CITATION STYLE
Ailon, N., Chien, S., & Dwork, C. (2006). On clusters in Markov chains. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 3887 LNCS, pp. 43–55). https://doi.org/10.1007/11682462_9
Mendeley helps you to discover research relevant for your work.