针对无序数据集异常检测的树突状细胞算法  

Dendritic cell algorithm for anomaly detection in unordered data set

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3期878-881,908,共5页Computer Engineering and Design

基　　金：国家自然科学基金项目(60975031)

摘　　要：为了提高树突状细胞算法对无序数据集的异常检测性能,分析了上下文环境的频繁转换是导致检测精度降低的主要原因,提出了一个"倍增-归并"的树突状细胞算法。先将数据集放大n倍,即每种抗原产生n个实例,对每个实例进行评估,综合每种抗原的n次评估得到最终结果。算法体现了细胞环境决定抗原状态的生物机制,通过倍增营造了相对稳定的环境,通过归并综合了多数正确判断减少了误判的影响。实验结果表明,该算法对无序数据集具有可观的检测精度和稳定的检测性能。To improve the anomaly detection performance of the dendritic cell algorithm （DCA） in unordered data sets, considering that with the context changing multiple times in quick succession there will be a sudden drop in accuracy, a multiplying and merging dendritic cell algorithm （MMDCA） is proposed. Firstly, the data set is multiplied n times, n instances are generated for each type of antigen, then each instance is finally the n assessments of each type of antigen is merged to get the final result. The algorithm implies the biological rhenism that the state of the antigen is determined by the context, multiplying will result in the relatively stable context, and merging can combine most correct judgments so as to reduce the influence of the errors Experiments show that the algorithm presented has considerable detection accuracy and stable detection performance in the unordered data set.

关 键 词：树突状细胞算法 无序数据集 异常检测 危险理论 数据融合 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]