Models Comparison Based On Intrusion Detection Using Machine Learning

Network security is the greatest challenges in our current generation. Network intrusion detection (NID) is one of the fundamental techniques used to protect computer networks from threads. However, there is contemplation on the possibility and sustainability of the traditional approaches employed especially with the current modern networks. Recently, majority of the researchers employ the application of machine learning models such as logistic regression (LR), random forest (RF), K-nearest neighbor (KNN) and Support Vector Machines (SVM) among others to address the NID problem. In this research, the Network intrusion classification system has been analyzed based on KDD Cup ’99 data set. In the classifier implementation section, the three most widely used models were employed; K-nearest neighbor, random forest and logistic regression as our classification algorithms. The attack detection accuracy and training time was used to evaluate the models. The random forest provides the best detection accuracy of 99.5% and 74 second training time provide by logistic regression