【摘要】 在网络安全问题日益突出的今天,如何迅速而有效地利用入侵检测系统发现各种入侵行为,对于保证系统和网络资源的安全十分重要。改进的k-means聚类算法解决了传统聚类算法在入侵检测领域所面临的混合类型数据相异度计算的问题。理论分析表明,此方法具有较好的时间复杂度,适合采用增量聚类,具有较好的扩展性,而且适用于任何数据类型,可应用于大规模的数据集。更多还原

【Abstract】 Today,network security issues are increasingly prominent,how quickly and effectively using the detection system find out various activities to ensure the safety of the system and network resources is important.Improved k-means clustering algorithm solves the problems of the conventional clustering algorithms in the field of intrusion detection.Theoretical analysis showes that this method has a good time complexity for the incremental clustering,has etter extensibility,and could be applied to any data types and mass of data collection.更多还原