Support vector machines (SVMs) have been successfully employed to solve NL regression and time series problems. The possibility of extending the application of SVMs to software reliability forecasting is investigated through the use of simulated annealing algorithms to select the parameters of an SVM model. Numerical examples are used from existing literature to demonstrate the performance of software reliability forecasting. The results indicate better performance of the SVMSAs and validity of the model as an alternative for forecasting software reliability.
展开▼