Enhancing Software Defect Prediction Through Advanced Machine Learning: Investigating Solutions to Key Limitations of Traditional Techniques

Software defect prediction (SDP) ensures high-quality and reliable software systems. However, traditional SDP techniques face various challenges that hinder their effectiveness. This study investigates the key challenges of traditional SDP approaches and proposes promising solutions to overcome these limitations. The paper examines the issues of data availability, feature selection, imbalanced data, and lack of contextual information in traditional techniques. It presents innovative solutions, including advanced machine learning (ML) techniques, feature engineering, context-aware approaches, hybrid models, and continuous monitoring and feedback. To evaluate the performance of SDP, the simulation results are compared on four datasets, PC1, CM1, JM1, and PC3, for Random Forest (RF), Support Vector Machine (SVM), K-nearest neighbors (KNN), Logistic Regression (LR), Linear Discriminant Analysis (LDA), Decision Tree (DT), and Naive Bayes (NB) models. Experimental implementations using ML-based techniques with feature reduction are conducted to validate the proposed solutions. The goal is to enhance SDP's accuracy, relevance, and efficiency, ultimately improving software quality assurance processes.