A software reliability growth model for open-source software using sine cosine algorithm

Software reliability models are used to predict the trustworthiness of software systems over time, based on the feedback obtained from testing efforts. From existing literature, a few models have been recognized for open-source software. In this paper, a software reliability growth model for open-source software has been discussed that incorporates generalized extended inverse Weibull distribution as a testing effort function. The sine–cosine algorithm has been used to estimate unknown parameters. The proposed work has been justified with numerical examples using real software failure data. For the comparative study, statistical methods, i.e., mean square error, R-squared, Theil statistic (TS), and graphical representation of the results have been used. The outcomes of the proposed model show better goodness of fit and predict significantly compared to the existing models. The suggested study may be extended by considering uncertainty factor in the future.