A STUDY ON OPERATIONAL SOFTWARE RELIABILITY PREDICTION BASED ON DEVELOPMENT PROJECT DATA

This study discusses prediction methods of software failure occurrences which are caused after software development. We analyze the actual data sets collected from the real software development companies by applying several regression methods and a neural network model. The data sets consist of quantitative data and qualitative ones. We compare the prediction performance of the operational software reliability by using these methods and data sets. In particular, we consider whether qualitative data sets can be used for the software reliability prediction