Development of Probabilistic-Logic Inference in Algebraic Bayesian Networks

Abstract

The paper presents a brief history of the probabilistic graphic models formation and examines the reasons for their appearance. Special attention is given to Bayesian trust networks and related algebraic Bayesian networks, which are a representation of knowledge bases with uncertainty. Key achievements in the development of probabilistic-logic appartus in algebraic Bayesian networks and Bayesian trust networks are presented along with examples of the Bayesian trust networks using cases in software products. The review of the probabilistic-logic inference apparatus development is continued with the matrix-vector formulation and proof of the solution of the first task posterior inference. In addition, the problems in algebraic Bayesian networks currently being faced by researchers are listed. This paper might be useful to lecturers of disciplines on modern research in the field of artificial intelligence, to students specializing in information technologies, who are considering promising topics for their term papers, final or research papers and to the staff of it companies, exploring the possibilities of applying mathematical models in business processes.