Hybrid Model to Address Class Imbalance Problems in Software Defect Prediction using Advanced Computing Technique

One of the most well-known study areas in computer science is software defect prediction. It aims to find defects occurring from the method level code, so it can be used to better prioritize software quality assurance work. This research work has gone through different methods for predicting the software defect from the code, applied different class balancing techniques for imbalanced class data, and removed multi-collinearity from the data which helped in giving better results. The best trained model can predict the defect with 92% Accuracy, 96% Recall, 89% Precision, and a 0.92 ROC AUC Score.