Modeling Reliability Growth among Different Issue Types for Multi-Version Open Source Software

This paper aims to propose a quantified approach for a multiple versions of software product released based on the implementation of three different issue types. The issue types being bug fixing, new feature introduction and feature improvements. Multiple releases helps in providing a more reliable software and also requires managing these software releases. The models provide a quantitative approach to predict the failure that may occur for modeling reliability growth based on three parameters. We are considering three criteria for goodness-of-fit, such as MSE, RMSE and R
²
for performance comparison of models. The analysis with hive data set shows that model2 fits considerably better. Our approach will be helpful in planning the release of the software. A reliable software with proper release planning is the need of the hour in our ever changing technological world.