Software-Defect-Prediction

##Objective: Most of the existing software defect prediction techniques have roughly to do with software metrics and classification algorithms. We develop the software defects classification prediction based on the software development repository by adopting the popular software development source techniques. We can get the data from NASA-mdp datasets, and then analyze the characters of these data sets use different models created on various characters, indicate others. Still, suitable software metrics that can build different prediction models. And then, We can use these models in a future version of the software, predict software defects on different granularity, and then compare other models' performance. Thoroughly know the importance of software defect prediction technology, giving to existing methods of software defect classification forecast, coming up with one of the critical technologies, counting metrics collection selected, classifier technology used to construct different models, and data set preprocessing and performance evaluation.