This article is cited in 8 papers

Measures of truth and probabilistic graphical models for representation of knowledge with uncertainty

A. A. Fil'chenkov<sup>ab

a</sup> St. Petersburg Institute for Informatics and Automation of RAS
^b St. Petersburg State University, Department of Mathematics and Mechanics

Abstract: For representation of knowledge with uncertainty both a mathematical formalism allowing to describe and handle uncertainty, and a theoretical computer model, limiting the requirement for such representation and processing of memory and time, are required. The paper gives an overview of the main truth measures including probability measure that being applied in artificial intelligence to represent uncertainty, and probabilistic graphical models which allow to limit the growth of processing algorithms complicity and memory requirements for representation of knowledge with uncertainty by means of computation localisation.

Keywords: Bayesian networks, internal and external measures, probabilistical graphic models, knowledge with uncertainty, measures of truth.

Received: 10.12.2012

Bibliographic databases:

•