Intrusion detection based on PCA and fuzzy clustering optimized by CS
Zixuan Li, Yixin Su, Qihang Han · 2017
Considering the high dimension of network intrusion data samples and the sensitivity to the selection of initial cluster centers of the traditional Fuzzy C-Means clustering algorithm (FCM) which is easy to be trapped around the local optima, a new intrusion detection algorithm is presented based on PCA and Fuzzy C-Means clustering optimized by CS in this paper. First, we extract the features of data by PCA and eliminate the redundant attributes among the data. Then, we use the improved algorithm, whose selection of initial cluster centers is optimized by cuckoo search so that not to fall into local optimal solution easily, to cluster the data samples. Simulation contrast experiments on the data sets of KDDCUP (1999) show that this algorithm can reduce the computation of intrusion detection with a higher detection rate and lower false positive rate.