References & Citations
Computer Science > Formal Languages and Automata Theory
Title: Profinite Techniques for Probabilistic Automata and the Markov Monoid Algorithm
(Submitted on 13 Jan 2015 (v1), last revised 9 Sep 2017 (this version, v4))
Abstract: We consider the value 1 problem for probabilistic automata over finite words: it asks whether a given probabilistic automaton accepts words with probability arbitrarily close to 1. This problem is known to be undecidable. However, different algorithms have been proposed to partially solve it; it has been recently shown that the Markov Monoid algorithm, based on algebra, is the most correct algorithm so far. The first contribution of this paper is to give a characterisation of the Markov Monoid algorithm. The second contribution is to develop a profinite theory for probabilistic automata, called the prostochastic theory. This new framework gives a topological account of the value 1 problem, which in this context is cast as an emptiness problem. The above characterisation is reformulated using the prostochastic theory, allowing us to give a simple and modular proof.
Submission history
From: Nathanaël Fijalkow [view email][v1] Tue, 13 Jan 2015 13:28:30 GMT (53kb,D)
[v2] Fri, 30 Jan 2015 08:35:06 GMT (53kb,D)
[v3] Fri, 12 Feb 2016 09:38:52 GMT (469kb)
[v4] Sat, 9 Sep 2017 18:59:16 GMT (175kb,D)
Link back to: arXiv, form interface, contact.