Integrasi SMOTE pada Naive Bayes dan Logistic Regression Berbasis Particle Swarm Optimization untuk Prediksi Cacat Perangkat Lunak

Software defects are one of the main contributors to information technology waste and lead rework, thus consuming a lot time money. defect prediction has objective prevention by classifying certain modules as defective or not defective. Many researchers have conducted research in field software using NASA MDP public datasets, but these datasets still shortcomings such class imbalance noise attribute. The problem can be overcome utilizing SMOTE (Synthetic Minority Over-sampling Technique) attribute solved selecting features Particle Swarm Optimization (PSO), So this research, integration between PSO is applied classification technique machine learning naïve Bayes logistic regression. From results experiments that been carried out on 8 dividing dataset into training testing data, it found + each improve performance with highest AUC (Area Under Curve) value average 0,89 regression 0,86 at same better than without combining two.