一种事件序列相似性评估方法

An Estimate Method of Similarity of Event Sequences

事件序列广泛出现在工业制造和信息科学的各个领域,如流水线中的操作流、字符串、用户访问网站的点击流、DNA序列、系统维护记录等。不同于数字信号处理中的时间序列,事件序列的自变量是代表顺序的有限正整数,因变量是表征事件的常量,这样的常量没有大小之分,是定性的特征。文章立足于实际应用,通过把问题转化为一个线性规划问题,建立了一种局部特征适应性好的相似性评估方法,并成功应用于某型导弹武器系统的考核评估中。

The event sequences appear in various fields of industrial manufacturing and information science widely, such as the operating flow of the pipeline＇s, string, click-stream came from users access the site, DNA sequences, system maintenance records. Unlike digital sig nal processing in the time sequence, the independent variables of the event sequences are the representative sequences of finite positive inte ger. And the dependent variables are the characterization of the constants of the events which are no size differences. Based on the practical application, the problem into is converted a linear programming problem, and a local good adaptability similarity assessment methods is es tablished, The methods applied in the examination and evaluation of a certain type of missile weapon system successfully.