An Improved Ant-based Classifier for Intrusion Detection

  • Authors:
  • Junbing He;Dongyang Long;Chuan Chen

  • Affiliations:
  • Sun Yat-sen University, China;Sun Yat-sen University, China;Sun Yat-sen University, China

  • Venue:
  • ICNC '07 Proceedings of the Third International Conference on Natural Computation - Volume 04
  • Year:
  • 2007

Quantified Score

Hi-index 0.00

Visualization

Abstract

Ant based classifier has been proposed to extract classification rules that can predict the class label of an unlabeled instance, but in the field of intrusion detection it is relatively unexplored. In this paper we describe a variety of modifications that we have made to the data mining algorithms in order to improve accuracy and efficiency. We also implement the modified algorithm on intrusion detection. The ant colony algorithm is employed to derive a set of classification rules from network audit data. Experiment result and comparative study shows our approach is effective and practical.