Intrusion Detection Based on Feature Transform Using Neural Network

摘要: In this paper, a novel method for intrusion detection is presented. The presented uses clustering based on the transformed features, which can enhance effectiveness of clustering. Clustering used in anomaly systems to separate attack and normal samples. general, separating samples original input space not an easy task. For better separation samples, transformation that maps data into different feature should be performed. we propose obtaining proper function reflects characteristics given domain. obtained from hidden layer trained three-layer neural network. Experiments over network connection records KDD CUP 1999 set are evaluate proposed method. result experiment clearly shows outstanding performance