Confidence Interval for a Software Reliability Model with a Bathtub-Shaped Fault Detection Rate
Dae Kyung Kim · Journal of Applied Reliability · 2020
Purpose: In recently years, a number of software reliability models (SRMs) have been developed by researchers. In this study, we were interested in the confidence interval of the mean value function (MVF) for SRM with bathtub-shaped fault detection rates (FDRs) model. The confidence interval was considered because point estimates are very unstable when the sample size is small, such as in the initial test phase [1, 2].BRMethods: In our research, we calculated the confidence interval of the MVF. First, the MLE was calculated using Obha data [8], and then the confidence interval of the MVF was calculated using Fisher’s information matrix.BRResults: We obtained the confidence interval for the MVF using the MLE and Fisher’s information matrix.BRConclusion: In this paper, the confidence interval of the MVF was obtained for a model in which the software has a bathtub-shaped failure rate. This result can be applied to models with little failure data during the test period in order to estimate stable reliability measures.