Abstract
Generally a software reliability growth model (SRGM) through a non-homogenous Poisson process (NHPP) is represented by its mean value function originated from the cumulative distribution function of a positive-valued continuous random variable. We consider some truncated distributions of positive-valued random variables and use them in the mean value function of NHPP. The chief characteristics of the resulting SRGMs are presented. Maximum likelihood estimation of parameters is discussed. Reliability aspects of live data sets are evaluated. Relative comparison among the models is also studied.
Get full access to this article
View all access options for this article.