EUVIMEDEUVIMED entdecken
Crossref

A Comparative Evaluation of Random Forest and XGBoost Models for Disease Detection Using Medical Indicators

Murteza Hanoon Tuama

International Journal of Professional Studies · 2025 · 19 (1) · 11-18

DOI: 10.37648/ijps.v19i01.002 ↗

Kurz gefasst

Fachartikel in der Zeitschrift International Journal of Professional Studies, 2025. Verfasst von Murteza Hanoon Tuama. Die Arbeit wurde bisher 3-mal zitiert (Crossref).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

In the field of 21st-century medicine, Machine Learning (ML) has become the cornerstone of healthcare by transforming disease detection at an earlier stage by facilitating early-stage accurate examination of medical attributes of the individuals including cholesterol, blood pressure and heart rate levels. Here we present an added comparison of two of the stratum of ensemble learning learned models: Random Forest and XGBoost on real-life medical datasets. The performance of the models was compared using Accuracy, Precision, Recall, F1-Score, and ROC-AUC. Random Forest showed great stability with the best accuracy of 96.75%(0.967543) produced by its bagging technique to avoid overfitting. Using the XGBoost model, an accuracy of 96.45% (0.964451) was achieved; this is possible, to an extent due to the ability for XGBoost to handle imbalanced datasets due to its (arguably) superior gradient boosting technique. These algorithms were very powerful showing up to a 20% reduction in diagnosis errors and a 30% reduction in diagnosis time compared with traditional methods but chronic diseases like cardiovascular diseases and diabetes were the most benefited by this type of models. It distinguishes itself through the weight it gives to bringing these models out of the lab and into real practice, which should illuminate their benefits across a constellation of ranges, from improvement in diagnostic pathways to pathology, to reduced costs for niche resources, to streamlined workflows in decision-making. For example, they can facilitate identification of high-risk patients early in procedures, which allows intervention to be focused on this cohort and a subsequently better outcome for the patient. In addition, this study addresses all primary challenges & data quality, model's adaptability to the demographics, and ethical issues that will ensure fairness and transparency in ML dependent actions. Pioneering tools will make healthcare efficient, equitable, and accessible across the globe by linking theory with practical application.

Zitieren

APA

Tuama, M. H. (2025). A Comparative Evaluation of Random Forest and XGBoost Models for Disease Detection Using Medical Indicators. International Journal of Professional Studies, 19(1), 11-18. https://doi.org/10.37648/ijps.v19i01.002

Vancouver

Tuama MH. A Comparative Evaluation of Random Forest and XGBoost Models for Disease Detection Using Medical Indicators. International Journal of Professional Studies. 2025;19(1):11-18. doi:10.37648/ijps.v19i01.002

BibTeX
@article{Tuama2025,
  author = {Tuama, Murteza Hanoon},
  title = {A Comparative Evaluation of Random Forest and XGBoost Models for Disease Detection Using Medical Indicators},
  journal = {International Journal of Professional Studies},
  year = {2025},
  volume = {19},
  number = {1},
  pages = {11--18},
  doi = {10.37648/ijps.v19i01.002}
}
Kostenlos starten →