Deriving belief networks and belief rules from data: a progress report

  • Authors:
  • Jerzy W. Grzymała-Busse;Zdzisław S. Hippe;Teresa Mroczek

  • Affiliations:
  • University of Information Technology and Management, Rzeszów, Poland and Department of Electrical Engineering and Computer Science, University of Kansas, Lawrence, KS;University of Information Technology and Management, Rzeszów, Poland;University of Information Technology and Management, Rzeszów, Poland

  • Venue:
  • Transactions on rough sets VII
  • Year:
  • 2007

Quantified Score

Hi-index 0.00

Visualization

Abstract

An in-house developed computer program Belief -SEEKER, capable to generate belief networks and also to generate sets of belief rules, has been presented in this paper. This system has a modular architecture, and consists of the following modules: Knowledge Discovery Module (KDM, an intelligent agent or pre-processor), Belief Network Development Module (BDM, generates belief networks), Belief Network Training Module (BTM, shows the distribution of conditional probabilities using a two-dimensional graph, together with some hints extracted from the investigated data), Belief Network Conversion Module (BCM, converts generated belief networks into relevant sets of belief rules of the type IF...THEN), and Probability Reasoning Module (PRM, checks the correctness of developed learning models as the "prediction of future" in classification of unseen examples).