It can be useful to ensure software quality as it tries to predict defects of the modules of the software.
In this project some supervised machine learning techniques were used including ANN, CNN, SVM, XGBoost, Decision Tree, Random Forest to train a robust model to predict defects of the softwares. After fetching the results from the above techniques, we combine them through Logistic Regression and get the final output.
Open source datasets including JM1, CM1, AR6, KC1, MC1, PC2 from NASA PROMISE Data Repository to perform this comparative study.
Widely used metrics for binary classification task were used such as accuracy, precision, recall, F1 score, AUC ROC. Result tables are included also.