TL;DR
The results indicate that traditional ML models, especially random forest and extra trees, are still very effective for metric-based defect prediction, while DL and multi-modal approaches need to be fed with richer software artifacts to reach their full potential.
Class imbalance can make software defect predictors appear successful while missing defective modules. This study compared no correction, random oversampling, random undersampling, SMOTE, ADASYN, and class-weighted learning across logistic regression, decision tree, random forest, support vector machine, and neural network classifiers. KC1 and PC1 NASA/PROMISE datasets (3,218 modules; 403 defective) were evaluated by stratified five-fold cross-validation. Imputation, scaling, and correction were fitted only within training folds. Precision, sensitivity, specificity, F1-score, balanced accuracy, ROCAUC, PR-AUC, and confusion matrices were reported. Across classifiers, baseline balanced accuracy was 0.580; corrected means ranged from 0.696 to 0.720. Correction increased sensitivity but generally reduced precision and specificity. ROS achieved the highest mean F1-score (0.388), while SMOTE achieved the highest mean PR-AUC (0.378). A Friedman comparison indicated heterogeneity among techniques, followed by Holm-adjusted paired Wilcoxon tests. No approach dominated every classifier or metric. Leakage-safe correction and multi-metric assessment are essential; accuracy alone is unsuitable for selecting defect predictors.
An intelligent machine learning-based bug prediction framework that uses SMOTE for dataset balancing and feature selection to identify the most relevant software metrics and uses advanced ensemble learning techniques, such as CatBoost, LightGBM, and the Stacking Ensemble model, to improve prediction accuracy.
Software reliability remains a major concern in modern software engineering due to the increasing complexity of software systems and the rapid pace of development. Software Failure Prediction (SFP) aims to identify fault-prone modules before deployment, enabling organizations to reduce maintenance costs and improve system quality. Artificial Intelligence (AI) and Machine Learning (ML) techniques provide data-driven approaches for analyzing software metrics, defect repositories, and execution logs to predict failures. This review examines AI/ML-based approaches for software failure prediction and efficiency optimization, with explicit attention to benchmark datasets, dataset quality, preprocessing, class imbalance, feature selection, model families, validation strategies, and evaluation measures. Representative NASA and PROMISE/Jureczko datasets are characterized in terms of software-unit type, metric families, binary defect labels, and imbalance. The review emphasizes Precision, Recall, F1-score, ROC-AUC, and MCC in addition to accuracy and compares traditional ML, ensemble, and deep-learning approaches. It further discusses missing data, concept drift, and cross-project prediction as key factors affecting real-world generalization. The analysis indicates that no single model is universally optimal; robust software failure prediction requires dataset-aware preprocessing, leakage-safe validation, imbalance-aware evaluation, and an explicit trade-off among predictive performance, computational efficiency, and interpretability.
This study investigates the use of ensemble learning with Large Language Models (LLMs) to improve the accuracy of software vulnerability prediction, following a structured experimental approach to assess whether combining multiple models can enhance performance. Three baseline models, CodeBERT, GraphCodeBERT, and CodeT5, were trained and assessed on the Devign dataset, which provides a large collection of labeled source code snippets. Their outputs were then integrated using three ensemble techniques: Majority Voting, Weighted Voting, and Stacking. Precision, recall, and F1-score metrics were used to gauge performance. Ensemble approaches outperformed all standalone models. In particular, Majority Voting increased precision from 0.601 (CodeBERT) to 0.690, representing a 14.81% improvement. Keeping in view the detection accuracy, this study focused on reducing the false positives. The results show that the ensemble techniques are a practical approach to boost the precision of LLMs in the detection of vulnerabilities. Ensemble learning can address the challenges faced by standalone models by reducing false positives and improving the overall trade-off between accuracy and reliability. The study suggests that ensemble methods offer great potential in the advancement of software security analysis.
Software defect prediction is essential for maintaining code quality in critical domains, yet it remains challenging due to feature redundancy and class imbalance. This study proposes an optimized Pi–Sigma Neural Network (PSNN) framework leveraging Correlation-Based Feature Selection (CBFS) and Min-Max normalization. Utilizing the NASA PROMISE CM1 dataset, a 5-fold stratified cross-validation pipeline was implemented to ensure statistical robustness and prevent data leakage. Experimental results on the CM1 dataset show the refined PSNN achieves high performance (99.80% accuracy on CM1) after aggressive feature reduction to 3–5 features and a precision of 1.000. To address class imbalance, the model achieved a Matthews Correlation Coefficient (MCC) of 0.988 and a G-Mean of 0.990. Comparative analysis shows that the developed PSNN-FS, despite its simplicity, achieves strong performance competitive with more complex architectures on the CM1 dataset.
Abstract - Software defect detection is a crucial area in software engineering focused on identifying issues within software systems. To achieve software success, it is essential to bridge the gap between software engineering and data mining. Various methods, including clustering, statistical techniques, matrix-based neural networks, black box & white box testing, & machine learning, remain employed toward predict software defects. These methods are applied to enhance accuracy in defect prediction through machine learning. Objective of this study is toward improving software defect prediction five publicly available NASA datasets: CM1, JM1, KC2, KC1, and PC1. Feature assortment performances are cohesive through ML methods, such as Random Forest to achieve higher defect prediction accuracy compared to methods without feature selection (WOFS) and to get matrix problem we use Classification Network. A defect is defined as an imperfection caused by an error, fault, or failure in the software development process. In this context, an "error" refers to human actions leading to undesirable outcomes, while a "defect" denotes a decision resulting in incorrect outcomes when attempting to resolve a problem. Software defect prediction has gained prominence in recent years due to its direct impact on software quality. Defective software modules can affect product quality, leading to cost overruns, delays in project timelines, and increased maintenance expenses. Key Words: software defect prediction five publicly available NASA datasets: CM1, JM1, KC2, KC1, and PC1.
We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.
