Authors : Tae-Jin Yang and Jea-Gun Park
Abstract: Software testing (debugging) to reduce testing costs is essential for the software reliability. In this study, the comparative problem for the reliability model using the Weibull extension distribution based on the generalized order statistics and the flexible Weibull extension distribution which made out efficiency about the software reliability was proposed. The maximum likelihood estimation and bisection iteration method for the estimating parameters were used. In addition, the model selection based on the mean square error and coefficient of the determination for the efficient model were offered. Analysis of the failure time based on the Weibull extension distribution and the flexible Weibull extension distribution was employed for the proposing reliability model. The Laplace trend test was employed for the assurance property about the failure time. The result of this study was obtained that the proposed Weibull extension distribution model than the flexible Weibull extension model is more efficient in terms of the reliability under a fixed shape parameter condition. Thus, the Weibull extension distribution model can be used as a reliability alternative model. From this study, the software developers must be filled to the growth model by a prior knowledge for the software to identify the failure modes which can be benefited.
Tae-Jin Yang and Jea-Gun Park, 2016. A Comparative Study of the Software NHPP Based on Weibull Extension Distribution and Flexible Weibull Extension Distribution. International Journal of Soft Computing, 11: 276-281.