A Software Quality Measurement using Generalized Half Logistic Distribution

Abstract

Two types of generalizations are proposed for the Half logistic distribution by the authors. In this article an attempt is made to generate a software quality measurement with reliability growth models for the two types of generalizations to half logistic distribution and to compare the models. The models are developed using NHPP. The constants of the model are obtained using well known MLE. The Reliability Growth Models developed with GHLD is compared with generalized exponential distribution through goodness of fit.

Keywords: SDLC, software reliability, Non-homogeneous Poisson process (NHPP),    

                  Generalized half-logistic model, maximum likelihood estimation, Q-Q Plot.