Multiple Instance Learning with Multiple Objective Genetic Programming for Web Mining

  • Authors:
  • Amelia Zafra;Eva L. Gibaja;Sebastián Ventura

  • Affiliations:
  • Department of Computer Science and Numerical Analysis, University of Cordoba, Spain;Department of Computer Science and Numerical Analysis, University of Cordoba, Spain;Department of Computer Science and Numerical Analysis, University of Cordoba, Spain

  • Venue:
  • Applied Soft Computing
  • Year:
  • 2011

Quantified Score

Hi-index 0.00

Visualization

Abstract

This paper introduces a multi-objective grammar based genetic programming algorithm, MOG3P-MI, to solve a Web Mining problem from the perspective of multiple instance learning. This algorithm is evaluated and compared to other algorithms that were previously used to solve this problem. Computational experiments show that the MOG3P-MI algorithm obtains the best results, adds comprehensibility and clarity to the knowledge discovery process and overcomes the main drawbacks of previous techniques obtaining solutions which maintain a balance between conflicting measurements like sensitivity and specificity.