Proposed Performance Evaluation Of Some Selected Machine Learning Algorithms for Software Defect Prediction System Using Stochastic Gradient Decent Algorithm

Abstract

Software defects are critical issues that can compromise the quality, increase costs, and delay the development of software systems. This research evaluates the performance of various machine learning algorithms—Naive Bayes (NB), Support Vector Machine (SVM), Artificial Neural Network (ANN), K-Nearest Neighbor (KNN), and Decision Tree (C4.5)—for software defect prediction. The study employs the Stochastic Gradient Descent Algorithm (SGDA) to extract relevant features from publicly available datasets, ensuring the elimination of overfitting. These extracted features are then classified using the selected machine learning algorithms, with their performance evaluated based on accuracy, precision, sensitivity, and F1-score. The findings aim to identify the most effective classifier, thereby guiding the development of more reliable software defect prediction systems.