AnalyzingNB, DT and NBTree Intrusion Detection Algorithms

Abstract

This work implements data mining techniques for analysing the performance of Naive Bayes, C4.5 Decision Tree, and the hybrid of these two algorithms the Naive Bayes Tree (NBTree). The goal is to select the most efficient algorithm to build a network intrusion detection system (NIDS). For our experimental analysis we used the new NSL-KDD dataset, which is a modified dataset of the KDDCup 1999 intrusion detection benchmark dataset, with a split of 66.0% for the training set and the remainder for the testing set. In the testing process Weka has been used, which is a Java based open source framework consisting of a collection of machine learning algorithms for data mining applications. In terms of accuracy the experimental results show that the hybrid NBTree is more precise than the other two approaches and the decision tree is better than the Naive Bayes algorithm. Otherwise, in terms of speed of response the Naive Bayes outperform the other two algorithms followed by Decision Tree and NBTree, respectively.