摘要
针对无监督学习的入侵检测算法准确度不高、监督学习的入侵检测算法训练样本难以获取的问题,提出了一种粒子群改进的K均值半监督入侵检测算法,利用少量的标记数据生成正确样本模型来指导大量的未标记数据聚类,对聚类后仍未能标记的数据采用粒群优化的K均值聚类,有效提高分类器的分类准确性,并实现了对新类型攻击的检测。实验结果表明,算法的整体检测效果明显优于基于无监督学习和监督学习的检测算法。
An anomaly intrusion detection algorithm based on semi-supervised clustering along with PSO K-means was presented. It could solve the problems of the low detection rate of the intrusion detection algorithms based on unsupervised learning, and the insufficiency of training samples of the intrusion detection algorithms based on supervised learning. The algorithm utilized minimal labeled data and lots of unlabeled data to improve its learning capability, and novelty detection could also be carried out. The experimental results manifest that the detection results of the algorithm outperforms both the one based on unsupervised learning remarkably and the one based on supervised learning.
出处
《计算机应用》
CSCD
北大核心
2008年第7期1781-1783,共3页
journal of Computer Applications
基金
湖南省自然科学基金资助项目(06JJ5106)
湖南省教育厅科学研究项目(06C841)
湘潭大学研究生创新基金资助项目
关键词
半监督聚类
入侵检测
粒群优化
K均值
semi-supervised clustering
intrusion detection
particle swarm optimization
K-means