JISE


  [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18]


Journal of Information Science and Engineering, Vol. 36 No. 2, pp. 293-308


Computational Intelligence Algorithms to Handle Dimensionality Reduction for Enhancing Intrusion Detection System


HUSAM IBRAHIEM ALSAADI1,2, RAFAH M. ALMUTTAIRI3,
OGUZ BAYAT1 AND OSMAN NURI UCANI1
1Faculty of Engineering
Altinbas University
Istanbul, 34676 Turkey

2Faculty of Basic Education
University of Mustansiriy
Baghdad, 10052 Iraq

3College of Information Technology
University of Babylon
Babylon, 51002 Iraq
E-mail: husam.alsaadi@ogr.altinbas.edu.tr; rafahmohammed@gmail.com;
oguz.bayat@altinbas.edu.tr; osman.ucan@altinbas.edu.tr


In this paper, propose to use computational intelligence models to improve intrusion detection system, the computational intelligence algorithms are used as preprocessing steps for selecting most significant features from network data. Two computational intelligence algorithms, namely Ant Colony Optimization (ACO) and Particle Swarm Optimization (PSO) are implemented to generate subset of relevant features. The computational intelligence approaches have been applied to optimize the classification of algorithms. The most significant features obtained from computational intelligence is fed into the classification algorithm. Novelty of this presents research of use computational intelligence algorithms namely Ant Colony Optimization (ACO) and Particle Swarm Optimization (PSO) for handling dimensionality reduction. The dimensionality reduction is obstructed time processing of classification algorithms. Three classification algorithms namely K-Nearest Neighbors (KNN), Support Vector Machine (SVM) and Naïve Bayes (NB) are implemented for intrusion detection system. Benchmark datasets, namely, KDD cup and NSL-KDD datasets are used to demonstrate and validate the performance of the proposed model for intrusion detection. From the empirical results, it is observed that the classification algorithm has improved the intrusion detection system with using computational intelligence algorithms. A comparative result analysis between the proposed model and different existing models is presented. It is concluded that the proposed model has outperformed of conventional models.


Keywords: computational intelligence algorithm, classification algorithms, intrusion detection system, support vector machine, K-nearest neighbors

  Retrieve PDF document (JISE_202002_09.pdf)