An Efficient Intrusion Detection System on Various Datasets Using Machine Learning Techniques
Debraj Chatterjee · 2021
The sudden surge in the relentless usage of networks in the last few years has compelled the creation of Intrusion Detection Systems (IDSs), which are distinctively used to identify various irregularities and outbreaks persisting in these networks. To eradicate the existing problems of IDSs, various machine learning and data mining techniques are being employed to ease the detection process. This work aims to traverse through the various IDSs and some state-of-the-art machine learning classifiers that can be deployed to enhance the performance of the IDSs. The author also proposes a machine learning classifier–based model of an effective IDS that can detect a network good or bad with the help of different machine learning techniques. The Logistic Regression, Naïve Bayes, Stochastic Gradient Descent, K-Nearest Neighbours, Decision Tree, Random Forest, and Support Vector Machine classifiers have been implemented to calculate the accuracy of the models. Popular yet standard datasets KDD 1999, NSL KDD, and DARPA have been used for the implementation. The model gives a permissible accuracy of over 80% when tested for each of the attacks.