Abstract
Several change metrics and source code metrics have been introduced and proved to be effective features in building bug prediction models. Researchers performed comparative studies of bug prediction models built using the individual metrics as well as combination of these metrics. In this paper, we investigate whether the prediction accuracy of bug prediction models is improved by applying feature selection techniques. We explore if there is one algorithm amongst ten popular feature selection algorithms that consistently fares better than others across sixteen bench marked open source projects. We also study whether the metrics in best feature subset are consistent across projects.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
