Abstract
In this paper we propose a (sub)distribution-based bisimulation for labelled Markov processes and compare it with earlier definitions of state and event bisimulation, which both only compare states. In contrast to those state-based bisimulations, our distribution bisimulation is weaker, but corresponds more closely to linear properties. We construct a logic and a metric to describe our distribution bisimulation and discuss linearity, continuity and compositional properties.
Cite
CITATION STYLE
Yang, P., Jansen, D. N., & Zhang, L. (2017). Distribution-based bisimulation for labelled Markov processes. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 10419 LNCS, pp. 170–186). Springer Verlag. https://doi.org/10.1007/978-3-319-65765-3_10
Register to see more suggestions
Mendeley helps you to discover research relevant for your work.
