Software Defect Prediction Based on Optimized Machine Learning Models: A Comparative Study

Software defect prediction is crucial used for detecting possible defects in software before they manifest. While machine learning models have become more prevalent prediction, their effectiveness may vary based on the dataset and hyperparameters of model. Difficulties arise determining most suitable model, as well identifying prominent features that serve input to classifier. This research aims evaluate various traditional are optimized NASA MDP (Metrics Data Program) datasets. The datasets were classified using k-nearest neighbors (k-NN), decision trees, logistic regression, linear discriminant analysis (LDA), single hidden layer multilayer perceptron (SHL-MLP), Support Vector Machine (SVM). fine-tuned random search, feature dimensionality was decreased by utilizing principal component (PCA). synthetic minority oversampling technique (SMOTE) implemented oversample class order correct imbalance. k-NN found be several datasets, while SHL-MLP SVM also effective certain It noteworthy regression LDA did not perform other models. Moreover, outperform baseline terms classification accuracy. choice model should specific characteristics dataset. Furthermore, hyperparameter tuning can improve accuracy predicting defects.