Management of intrusion detection systems based-KDD99: Analysis with LDA and PCA

Abstract

Recently, the problem of the intrusion detection has been largely studied by the computer and networks security communities. Then, the Intrusion Detection System (IDS) becomes a interest topic in research and in particular in machine learning and data mining. In order to improve the classification accuracy and to reduce high false alarm rate from the classical data base like KDD99 or others. In this work, we present a state of the art about this topic and we use classification algorithms such as Linear discriminant analysis (LDA) and Principal Component Analysis (PCA) to identify the intrusion and classification anomaly. The experiments of the IDS are performed with NSL-KDD data set and we try to improve the existing classification methods.