On optimisation of the rough set-based decision rule extraction

  • Authors:

  • Urszula Stanczyk

  • Affiliations:

  • Silesian University of Technology, Institute of Informatics, Gliwice, Poland

  • Venue:
  • ACS'08 Proceedings of the 8th conference on Applied computer scince
  • Year:
  • 2008

Quantified Score

Hi-index 0.00

Visualization

Abstract

Extraction of decision rules that comprise a decision algorithm constitutes one of crucial steps within the rough set approach to the problem of classification. It follows the phase of finding all reducts of conditional attributes and for the one which is chosen the succeeding procedure of establishing all valid value reducts. From the variety of possible solutions there is required the one with the highest accuracy of classification as well as simplicity of implementation which is reflected by the lowest number of conditional clauses within the decision algorithm. In the paper there is described such optimising methodology employed to the rough set-based approach to the stylometric problem of authorship attribution.