Defect Prediction Using Akaike and Bayesian Information Criterion

Data available in software engineering for many applications contains variability and it is not possible to say which variable helps the process of prediction. Most work present defect prediction focused on selection best techniques. For this purpose, deep learning ensemble models have shown promising results. In contrast, there are very few researches that deals with cleaning training data parameter values from data. Sometimes high may cause a decrease model accuracy. To deal problem we used Akaike information criterion (AIC) Bayesian (BIC) variables train model. A simple ANN one input, output two hidden layers was instead complex AIC BIC calculated combination minimum be selected At first, were narrowed down smaller number using correlation values. Then subsets all combinations formed. end, an artificial neural network (ANN) trained each subset basis smallest value. It found only variables’ ns entropy as gives While, nm npt worst maximum