ARCHIVES
VOL. 2, ISSUE 4 (2026)
Performance analysis of bagging and boosting algorithms for diabetes prediction
Authors
Robin Sommer
Abstract
Diabetes mellitus is a major chronic disease that requires early
identification to support timely intervention and reduce the risk of long-term
complications. The increasing availability of electronic health records,
demographic information, lifestyle indicators, and clinical measurements has
created opportunities for machine learning-based diabetes prediction. Among
machine learning approaches, ensemble learning has attracted considerable
attention because it combines multiple predictive models to improve classification
performance and generalization. This article presents a performance-oriented
analysis of two important ensemble learning strategies, namely bagging and
boosting, for diabetes prediction. Bagging methods such as Random Forest
construct multiple decision trees independently using different bootstrap
samples, while boosting methods such as Gradient Boosting and Extreme Gradient
Boosting (XGBoost) sequentially develop weak learners to correct previous
prediction errors. Recent studies indicate that both approaches can effectively
model nonlinear relationships in diabetes-related data, although their
performance depends on data quality, class distribution, feature selection,
preprocessing, and validation methodology. The proposed analysis focuses on
comparing bagging and boosting according to accuracy, precision, recall,
F1-score, receiver operating characteristic area under the curve, computational
requirements, robustness, and interpretability. The study provides a
methodological framework for identifying the strengths and limitations of these
ensemble approaches and emphasizes the importance of evaluating models using
multiple performance measures rather than relying exclusively on accuracy.
Download
Pages:22-26
How to cite this article:
Robin Sommer "Performance analysis of bagging and boosting algorithms for diabetes prediction". World Journal of Applied Research, Vol 2, Issue 4, 2026, Pages 22-26
Pdf, Certificate, Cover Page, Index Page, Editorial Page, Accept Letter
Please enter the email address corresponding to this article submission.