EUVIMEDEUVIMED entdecken
Crossref

Enhancing Cross-Project Defect Prediction Using Stacking Ensemble and Hybrid Sampling

Aliyu Bashir Nuhu, Yusuf Salisu Ibrahim

International Journal of Innovative Science and Research Technology · 13.05.2026 · 123

DOI: 10.38124/ijisrt/26may202 ↗

Kurz gefasst

Fachartikel in der Zeitschrift International Journal of Innovative Science and Research Technology, online erschienen am 13.05.2026. Verfasst von Aliyu Bashir Nuhu und 1 weiteren Autor:innen. Die Arbeit nennt 28 Literaturangaben.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Software Defect Prediction known as SDP improves software quality by identifying defect-prone modules early in development, however, Cross-Project Defect Prediction (CPDP) remains challenging due to data heterogeneity across projects and severe class imbalance in defect datasets. Conventional machine learning models fail in effective generalization of new projects and demonstrate poor minority class detection in projects. This study aimed to develop and evaluate a Hybrid Ensemble Deep Learning framework to improve Cross Project Defect Prediction and performance under heterogeneous and imbalanced conditions. The framework combines Random Forest and XGBoost as base learners in a stacking generalization architecture, after which a Deep Neural Network serves as the meta-classifier. To address class imbalance and boundary noise, a SMOTE-Tomek hybrid sampling technique was joined into the preprocessing pipeline. The model was evaluated using a Leave-One-Project-Out (LOPO) validation approach on 5 different PROMISE datasets (CM1, JM1, KC1, MW1, and PC1).

Zitieren

APA

Nuhu, A. B., & Ibrahim, Y. S. (2026). Enhancing Cross-Project Defect Prediction Using Stacking Ensemble and Hybrid Sampling. International Journal of Innovative Science and Research Technology, 123. https://doi.org/10.38124/ijisrt/26may202

Vancouver

Nuhu AB, Ibrahim YS. Enhancing Cross-Project Defect Prediction Using Stacking Ensemble and Hybrid Sampling. International Journal of Innovative Science and Research Technology. 2026:123. doi:10.38124/ijisrt/26may202

BibTeX
@article{Nuhu2026,
  author = {Nuhu, Aliyu Bashir and Ibrahim, Yusuf Salisu},
  title = {Enhancing Cross-Project Defect Prediction Using Stacking Ensemble and Hybrid Sampling},
  journal = {International Journal of Innovative Science and Research Technology},
  year = {2026},
  pages = {123},
  doi = {10.38124/ijisrt/26may202}
}
Kostenlos starten →