Intrusion Detection System Based On Cluster

The raising constantly of the speed of the network. IDS must adapt to this speed, So IDS must improve speed about process data and decrease false positive and false negative, at the same time IDS be asked to find out unknown attack. Because use some knowledge about math and theory about IDS, this dissertation introduce used knowledge and talk about network security future. According this requirement, the dissertation introduce feature selection based on PCA as pretreatment of data, to selected data which has less dimension be processed with clustering that combine max-min value with C-average method. This method can reduce dimension of data and improve speed of data process. Clustering can distinguish new attack from already existed attack because clustering can cluster data according data character. Some customer may be changed their habit so clustering data must update their self with coming network data. This can reduce the number of training. In contrast to some method that already exist, from result about test it can draw conclusion that this method can reduce data process time and improve percent of detection at the same situation.