Intrusion detection based on multi-layer minimax probability machine classifier

The minimax probability machine (MPM) constructs a classifier, which provides a worst-case bound on the probability of misclassification of future data points based on reliable estimates of means and covariance matrices of the classes from the training data points, and achieves the comparative performance and the same complexity of time with a state-of-the-art classifier, the support vector machine. To improve the limited performance of minimax probability machine in training time, a new multi-layer classifier model based on MPM ensemble with boosting is proposed and is applied to intrusion detection in this paper. After illustrating our model with a representative dataset and applying it to the real-world network datasets KDD Cup 1999, simulations show that the intrusion detection system based on multi-layer minimax probability machine achieves the comparative performance with the support vector machine and with less training time.