An Analysis of Intrusion Detection Classification using Supervised Machine Learning Algorithms on NSL-KDD Dataset
Sarthak Rastogi, Archit Shrotriya, Mitul Kumar Singh, Raghu Vamsi Potukuchi
Abstract
From the past few years, Intrusion Detection Systems (IDS) are employed as a second line of defence and have shown to be a useful tool for enhancing security by detecting suspicious activity. Anomaly based intrusion detection is a type of intrusion detection system that identifies anomalies. Conventional IDS are less accurate in detecting anomalies because of the decision taking based on rules. The IDS with machine learning method improves the detection accuracy of the security attacks. To this end, this paper studies the classification analysis of intrusion detection using various supervised learning algorithms such as SVM, Naive Bayes, KNN, Random Forest, Logistic Regression and Decision tree on the NSL-KDD dataset. The findings reveal which method performed better in terms of accuracy and running time.