Abstract: Since the large size of data set in intrusion detection system, existing clustering algorithms can't give the answer in a proper time. This paper proposes a parallel K-means algorithm and evaluates the speedup ratio of it. Through the experiments it proves the efficiency and accuracy of this algorithm.

摘要： 由于入侵检测使用的数据集十分庞大,现有的串行聚类算法很难在合理的时间内得到结果。文章提出了一种应用于入侵检测的并行K-均值算法,给出了其加速比估算公式,实验证明了算法的正确性和有效性。

关键词: 入侵检测, 聚类, 未标记数据, K-均值

CLC Number: 

• TP393.08