非负矩阵分解在入侵检测中的应用

针对当前入侵检测系统存在的检测效果差,对训练数据集要求高的问题,提出了一种使用非负矩阵分解算法的异常入侵检测模型.在预处理阶段综合考虑系统调用数据的时序和频率特征,将进程的入侵检测问题转换为向量空间的异常点检测问题,利用非负矩阵分解在提取特征和数据降维方面的优点,将高维空间降维映射到低维空间,最终在低维向量空间实现入侵检测.实验证实本方法检测效果良好.

Application of Non-negative Matrix Factorization on Intrusion Detection

Current intrusion detection systems show poor performance on detection effect,and have strong restrict of training data sets.An anomaly detection model using non-netative matrix factorization is provided to solve those problems.Both sequence and frequency character are analyzed in the phase of preprocessing,and the problem of intrusion detection is converted to the problem of outlier detection of points in vector space.Excellence of non-negative matrix factorization in feature extracting and dimension reduce is made use of.And high dimension vector space is projected to low dimension space.Finally anomaly detection is achieved in low dimension space.Experiments are done and good performance is demonstrated.