Confidence Levels Assigned to Languages Generated under Imprecision
by
Mariana Neagu
Transilvania University of Brasov, Department of Computer Science
Coauthors: Cristina Vladarean (Transilvania University of Brasov, Department of Computer Science)

The paper presents some results concerning the representation of imprecision related to a family of languages generated by stochastic acceptors with imprecise internal behaviour. Based on previous results obtained for the application of fuzzy measures to databases, the present work is concerned with the definition of the confidence level as a pair of values corresponding to the belief and the plausibility measures.

Our considerations apply to the language generated by a model of stochastic automaton that describes with imprecision the transition from one internal state to another. The confidence level represents our guess about the next state of the automaton when reading one input symbol.

Using the analogy with the theory of the languages generated by stochastic automata, the present paper proposes some definitions for the selection predicate, for the cut-point, for the conjunctive and the disjunctive combinations of the confidence levels assigned to various input symbols. All these definitions lead us to the definition of the confidence level assigned to a sequence of input symbols.

Date received: March 13, 2000

Copyright © 2000 by the author(s). The author(s) of this document and the organizers of the conference have granted their consent to include this abstract in Atlas Conferences Inc. Document # caet-33.