中文摘要：

在网络入侵检测优化的研究中,对网络入侵特征进行准确检测,由于在复杂的网络环境中会存在大量噪声,传统的方法只是单一的入侵特征聚类方法,难以在包含大量噪声的复杂网络环境中进行入侵特征聚类。提出一种基于目标协同规划思想的网络入侵特征聚类方法。利用标准化处理过程和归一化处理过程对对网络入侵数据进行预处理,能够将原始的网络入侵特征属性映射到标准属性空间。提取入侵特征构成数据集合,并进行降维处理,为入侵特征的聚类提供了准确数据基础,将可能性模糊聚类算法和聚类中心分离的模糊聚类算法进行入侵特征聚类目标的协同规划,能够得到准确的聚类中心。实验结果表明,改进算法能够提高网络入侵聚类的准确率。

英文摘要：

A network intrusion feature clustering method based on the goal of collaborative planning is proposed. The network intrusion data are preprocessed with standardized process and normalization process, and the original network intrusion feature is mapped to the standard attribute space. Intrusion characteristics are extracted to consti- tute collection of data and dimension reduction is processed, so as to provide accurate data base for clustering of in- trusion features, probability fuzzy clustering algorithm and fuzzy clustering algorithm of cluster center separation are planned collaboratively for the goals of intrusion feature clustering, and accurate clustering center is acquired. The experimental results show that the improved algorithm can improve the accuracy of network intrusion clustering.