Networking and Communication Engineering

S. Ranjitha Kumari ¹, P. Krishna Kumari ²

Affiliations
1 School of Computer Studies (PG), RVS College of Arts and Science, Coimbatore, IN
2 Dept of MCA, Rathnavel Subramaniam College of Arts & Science, Coimbatore, IN

Abstract

Intrusion Detection System is gaining more popularity nowadays as everyone is keen on protecting their networks. It is used to identify various authentic and malicious activities in the system and network. A lot of research activities are taking place to protect the network from outsiders as well as insiders. Various soft computing techniques like Data Mining, Artificial Intelligence are used in Intrusion Detection System for identifying malicious activities. In this paper we have done a survey on four supervised machine learning algorithms: Decision Tree (J48), K-Nearest Neighbor (KNN), Naive Bayes (NB) and Support Vector Machine (SVM). We have shown a comparative analysis of these algorithms based on Accuracy, True Positive Rate (TPR) and False Positive Rate (FPR). We have used NSL-KDD dataset for our experiment. Based on the experimental result, we have shown that the performance of Decision Tree (J48) and K-Nearest Neighbor are better than other two algorithms in terms of Accuracy, True Positive Rate (TPR) and False Positive Rat (FPR).

Keywords

Intrusion Detection System, Machine Learning Algorithms, Naive Bayes (NB), Support Vector Machine (SVM), K Nearest Neighbor (KNN), Decision Tree (DT), NSL-KKD Dataset.

Full Text