A novel visual intrusion detection model based on projection pursuit

Abstract

Visualized information is an useful technique that can encode large amounts of complex interrelated data, being at the same time easily quantified, manipulated, and processed by a human user. This paper presents a novel visual intrusion detection classification model based on projection pursuit, which searches for the optimal projection direction by maximizing a projection index, projects network connection records from high-dimensional space into 1-dimensional space, and uses the projection values on the best direction to analyze the data structure and classify the type of intrusion. The processes and results of classification can be visualized by a scatter chart in cartesian co-ordinates. The experiments on KDD CUP 1999 dataset show that this model not only has good performance, but also makes the process of intrusion detection understandably.