关键词: 聚类, 网格, 密度, 入侵检测, 数据挖掘

Abstract: After discussing some problems in the current intrusion detection techniques, an intrusion detection method that applies cluster algorithm based on the density and grid was proposed. This algorithm shifted from non-dense part to dense part based on CLIQUE. Inaccuracy of clustering result of CLIQUE algorithm was avoided. The algorithm has the merit of grid-based clustering which is of low-complexity in time and space, and has the merit of density-based clustering which is of good noise immunity. Using the data sets of KDDCUP99, the results of simulation experiments show the effectiveness and feasibility of the clustering algorithm.

Key words: clustering, grid, density, intrusion detection, data mining