摘要
软件测试中收集的累积出错数据,往往是不完全的,使用它们对软件的可靠性进行分析,将影响到分析结果的精度.解决这一问题的途径有很多,本文试图应用 EM 算法于 NHPP(非齐次泊松过程)类模型的参数估计,以提高估计精度,从而提高软件可靠性分析的精确程度.
During software test phase,the accumulated failure data often are incomplete.Using them to anal- yse the software reliability will effect the result accuracy,there are many ways to solve this problem. In this article,we try to apply EM algorithm to estimating the values of the parameters of NHPP (Non-homogeneous Paisson)models,The estimating accuracy may be increased by this way.So the re- sult of software reliability analysis may be improved by this way too.
出处
《武汉大学学报（自然科学版）》
CSCD
1992年第4期17-22,共6页
Journal of Wuhan University(Natural Science Edition)
基金
国家自然科学基金
关键词
EM算法
G-O模型
非齐次泊松过程
条件概率密度
强度函数
EM algorithm
G-O model
non-homogeneous Paisson process
condition probability indensity
density function
