Comparative Software Infinite NHPP Reliability Growth Model Based on the Intensity Function of Log-Linear and Log-Power Forms
Tae-Jin Yang · The Smart Computing Review · 2014
Infinite failure non-homogenous Poisson process (NHPP) software reliability models presented in the literature exhibit either constant, monotonic increasing, or monotonic decreasing failure occurrence rates per fault. In this paper, comparison problems of the log-linear intensity function widely used in the field of reliability and the log-power intensity function of the reliability growth model are presented. The software failure situation was used in the infinite failure non-homogeneous Poisson process model. An algorithm to estimate the parameters was used for a maximum likelihood estimator and a bisection method. A model selection based on mean square error and coefficient of determination for the sake of an efficient model was employed. For analysis of software failure data, the log-linear model and log-power model were compared. In order to ensure the reliability of software failure data, a Laplace trend test was employed.