Bisimulation on Markov processes over arbitrary measurable spaces

3Citations
Citations of this article
3Readers
Mendeley users who have this article in their library.
Get full text

Abstract

We introduce a notion of bisimulation on labelled Markov Processes over generic measurable spaces in terms of arbitrary binary relations. Our notion of bisimulation is proven to coincide with the coalgebraic definition of Aczel and Mendler in terms of the Giry functor, which associates with a measurable space its collection of (sub)probability measures. This coalgebraic formulation allows one to relate the concepts of bisimulation and event bisimulation of Danos et al. (i.e., cocongruence) by means of a formal adjunction between the category of bisimulations and a (full sub)category of cocongruences, which gives new insights about the real categorical nature of their results. As a corollary, we obtain sufficient conditions under which state and event bisimilarity coincide. © 2014 Springer International Publishing Switzerland.

Cite

CITATION STYLE

APA

Bacci, G., Bacci, G., Larsen, K. G., & Mardare, R. (2014). Bisimulation on Markov processes over arbitrary measurable spaces. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 8464 LNCS, pp. 76–95). Springer Verlag. https://doi.org/10.1007/978-3-319-06880-0_4

Register to see more suggestions

Mendeley helps you to discover research relevant for your work.

Already have an account?

Save time finding and organizing research with Mendeley

Sign up for free