摘要
软件可靠性是软件质量的重要指标之一。传统的软件可靠性模型通过一些精确、无歧义的假设,对软件的可靠性增长行为进行抽象和简化。将未确知理论与聚类方法结合起来应用于软件可靠性的研究中,摆脱了对软件故障过程的各种分布假设,改善了模型应用中的不一致问题。最后通过实例对新算法和传统算法进行了比较与分析。
Software reliability is one of the important indexes of software quality. In traditional software reliability model, the behavior of software reliability growth was abstracted and simplified through some precise and unambiguous assumptions. Soft- ware reliability was researched by the way of integrating unascertained theory and clustering, discarding all kinds of assump- tions of distribution in software fault process and improving the incongruenee of model application to some extent. Finally, the new and traditional arithmetic were compared and analyzed with an example.
出处
《计算机应用研究》
CSCD
北大核心
2009年第9期3378-3380,共3页
Application Research of Computers
关键词
未确知理论
聚类
软件可靠性
模型
unaseertained theory
clustering
software reliability
model
