An Ensemble of Gaussian Mixture Model and Support Vector Machines for Network Intrusion Detection

NetworkIntrusion Detection Systems (NIDS) can protect computer networks and computersystems by detecting abnormal network packets taking agreed action plans,such as notifying an administrator or rejecting the packets. In this study,the aim is implementation of NIDS with improved performance using anensemble Support Vector Machines (SVMs) Gaussian Mixture Model(GMM). Four SVMs Radial Basis Function (RBF), linear, polynomial, andsigmoid kernel functions, a GMM were trained same portion withKnowledge Discovery Data Mining Tools Competition (KDD 99) dataset, andanother dataset was used to evaluate therespective models. Finally, five models integrated form ensembleIntrusion System (IDS) model test tovalidate its performance. The IDS SVM RBF function has thebest precision, recall, f1score, accuracy, false acceptance rate, rejection rate 99.88,99.67, 99.77, 99.82, 0.08, 0.33% respectively. ensemble built bycombining where each them equal voting rightsyields state-of-art performance, f1-score, falseacceptance 99.7, 99.4, 99.55, 99.65, 0.18 and0.59% respectively though it below SVM-RBF theSVM-polynomial Ensemble are expected have better performancethan single classifier, but result research shows that isnot applicable in all cases outperformed theensemble classifier.