|Communications on Applied Electronics
|Foundation of Computer Science (FCS), NY, USA
|Volume 6 - Number 7
|Year of Publication: 2017
|Authors: Neeraj Bisht, Amir Ahmad, A. K. Pant
Neeraj Bisht, Amir Ahmad, A. K. Pant . Analysis of Classifier Ensembles for Network Intrusion Detection Systems. Communications on Applied Electronics. 6, 7 ( Feb 2017), 47-53. DOI=10.5120/cae2017652516
In this paper an ensemble approach to classify network security data have been presented. The experiments are carried out with Decision trees and Naïve Bayes classifiers and ensemble them with methods like Bagging, Adaboost.M1, Random Forests, MultiBoosting, Rotation Forest and Random Sub Space on NSL KDD dataset which is a modified KDD anomaly detection dataset . Results with different performance measures suggest that no single classification method is the best for all types of datasets on all type of performance measures. The results based on the experiments have been tabulated and their comparative performance suggests that the decision trees ensembles performed better than the Naïve Bayes ensembles. Results also suggest that single decision tree is a good classifier for this data as it has reasonable classification accuracy and less training and testing time.