Inductive Learning with a Computational Network

  • Authors:
  • H. L. Viktor;I. Cloete

  • Affiliations:
  • Department of Informatics, University of Pretoria, Pretoria 0002, South Africa/ e-mail: hlviktor@econ.up.AC.ZA;Department of Computer Science, University of Stellenbosch, Stellenbosch 7602, South Africa.

  • Venue:
  • Journal of Intelligent and Robotic Systems
  • Year:
  • 1998

Quantified Score

Hi-index 0.00

Visualization

Abstract

This paper introduces a computational network which combinesheterogeneous rule-extraction algorithms for intelligent data analysis.Combining induction programs may alleviate the possible negative effects ofdata set representation and individual program’s influences, such asinductive bias. The application of the computational network to a diabetesdata set shows that, when combining the various programs, an increase inrule set accuracy and comprehensibility are obtained.