摘要
针对无监督的入侵检测检测效率较低,而有监督入侵检测算法不能有效的检测异常攻击,提出一种半监督学习的入侵检测算法,新算法先用有标记数据进行初始聚类,然后利用初始聚类指导未标记数据聚类,最后使用K近邻算法对仍没有确定类别的未标示数据对异常进行检测,结果表明,改进后算法的效果优于无监督和有监督学习的入侵检测算法。
In view of the low detection rate of unsupervised intrusion detection algorithm, and the inefficacy of the su- pervised intrusion detection algorithms for unknown attacker, in this paper, an semi-supervised learning based intru- sion detection algorithm is proposed. In this algorithm, a small number of labeled samples is used to obtain the initial clustering model, which is used to guide the unlabeled data clustering. The unmarked data whose categories are not determined are clustered with K-nearest neighbor algorithm to detect abnormalithy, experiment results show that the improved algorithm has better performance than the intrusion detection algorithms based on unsupervised and supervised learning.
出处
《成都信息工程学院学报》
2012年第6期560-563,共4页
Journal of Chengdu University of Information Technology
