用于实时异常检测的树突状细胞算法  

Dendritic cell algorithm for real-time anomaly detection

作　　者：袁嵩[1,2] 陈启卷[1] 

机构地区：[1]武汉大学动力与机械学院,湖北武汉430072 [2]武汉科技大学计算机科学与技术学院,湖北武汉430065

出　　处：《计算机工程与设计》2013年第1期159-162,共4页Computer Engineering and Design

基　　金：国家自然科学基金项目(60975031)

摘　　要：为了提高异常检测系统的实时性能,在尽可能短的时间内对出现的异常做出反应,对传统的树突状细胞算法的离线分析过程进行了改进,提出了一种实时分析算法。当抗原被足够多的树突状细胞提呈后立即输出评估结果,从而到达实时或接近实时分析的目的,足够的评判次数减少了误判的影响,与时间序列相关的抗原信号池消除了无关数据的相互干扰,和生物免疫相关的一些随机因素的考虑更加近似地模拟了生物系统的表现。实验结果表明,该算法在实时分析的基础上还具有可观的检测精度。To perform the anomaly detection in real time, continuously detect abnormal behaviors as soon as they occur, a real- time analysis algorithm is put forward to improve the offline analysis of the classical dendritic cell algorithm. When an antigen is presented by sufficient dendritic cells, it will be immediately assessed and output, thus the purpose of real-time or near-to real- time analysis can be achieved. Sufficient assessments can reduce the influence of the errors, the antigen and signal pool of tempo- ral correlation is designed to eliminate the mutual interference of the antigens and signals which are far apart, and the considera- tion of some random factors immune-inspired will make the algorithm much more similar to the biological system. The results of the experiments show that the real-time analysis algorithm proposed has the considerable detection accuracy.

关 键 词：人工免疫 危险理论 树突状细胞算法 异常检测 实时分析 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]