What's decidable about weighted automata?

71Citations
Citations of this article
19Readers
Mendeley users who have this article in their library.
Get full text

Abstract

Weighted automata map input words to numerical values. Applications of weighted automata include formal verification of quantitative properties, as well as text, speech, and image processing. In the 90's, Krob studied the decidability of problems on rational series, which strongly relate to weighted automata. In particular, it follows from Krob's results that the universality problem (that is, deciding whether the values of all words are below some threshold) is decidable for weighted automata with weights in ℕ ∪ {∞}, and that the equality problem is undecidable when the weights are in ℤ ∪ {∞}. In this paper we continue the study of the borders of decidability in weighted automata, describe alternative and direct proofs of the above results, and tighten them further. Unlike the proofs of Krob, which are algebraic in their nature, our proofs stay in the terrain of state machines, and the reduction is from the halting problem of a two-counter machine. This enables us to significantly simplify Krob's reasoning and strengthen the results to apply already to a very simple class of automata: all the states are accepting, there are no initial nor final weights, and all the weights are from the set {∈-∈1,0,1}. The fact we work directly with automata enables us to tighten also the decidability results and to show that the universality problem for weighted automata with weights in ℕ ∪ {∞}, and in fact even with weights in ℚ≥0 ∪ {∞}, is PSPACE-complete. Our results thus draw a sharper picture about the decidability of decision problems for weighted automata, in both the front of equality vs. universality and the front of the ℕ ∪ {∞} vs. the ℤ ∪ {∞} domains. © 2011 Springer-Verlag.

Cite

CITATION STYLE

APA

Almagor, S., Boker, U., & Kupferman, O. (2011). What’s decidable about weighted automata? In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 6996 LNCS, pp. 482–491). https://doi.org/10.1007/978-3-642-24372-1_37

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