EUVIMEDEUVIMED entdecken
Crossref

IMPLEMENTATION OF THE XGBOOST ALGORITHM FOR ACUTE RESPIRATORY INFECTION CLASSIFICATION

Putri Dwi Agnesya, Ani Dijah Rahajoe, Afina Lina Nurlaili

JOURNAL OF SCIENCE AND APPLIED ENGINEERING · 12.05.2026 · 9 (1) · 38-46

DOI: 10.31328/jsae.v9i1.7400 ↗

Kurz gefasst

Fachartikel in der Zeitschrift JOURNAL OF SCIENCE AND APPLIED ENGINEERING, online erschienen am 12.05.2026. Verfasst von Putri Dwi Agnesya und 2 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Acute Respiratory Infection (ARI) is a major public health concern in Indonesia and continues to contribute significantly to morbidity rates, particularly among children and vulnerable populations. Accurate clinical diagnosis is often challenging because several ARI subtypes share similar physiological manifestations, making differentiation based solely on visual or symptomatic examination highly prone to error. Conditions such as tonsillitis, acute bronchitis, acute nasopharyngitis, and pneumonia frequently exhibit overlapping symptoms, including cough, fever, shortness of breath, and sore throat. This diagnostic complexity highlights the need for computational decision-support systems capable of enhancing accuracy and efficiency in clinical environments, especially in primary healthcare facilities where time and diagnostic resources may be limited. In this study, a multi-class classification model was developed using the Extreme Gradient Boosting (XGBoost) algorithm, leveraging 21 binary-coded clinical symptoms extracted from 251 patient medical records across two public health centers. The dataset underwent preprocessing steps including symptom normalization, binary feature encoding, and implementation of two train–test split schemes (90:10 and 80:20) to assess model robustness. Model performance was evaluated based on accuracy, precision, recall, and F1-score to ensure comprehensive assessment across multiple metrics. The experimental results show that the XGBoost model optimized using GridSearchCV achieved the highest performance, recording an accuracy of 88% in the 90:10 split configuration with the optimal tuned parameters (learning_rate = 0.05, max_depth = 3, subsample = 1.0). These findings indicate that an optimized XGBoost classifier can serve as a reliable and effective decision-support tool for ARI subtype classification. Furthermore, the model demonstrates potential for integration into clinical workflows to assist medical practitioners in making faster and more accurate diagnostic decisions in primary healthcare settings.

Zitieren

APA

Agnesya, P. D., Rahajoe, A. D., & Nurlaili, A. L. (2026). IMPLEMENTATION OF THE XGBOOST ALGORITHM FOR ACUTE RESPIRATORY INFECTION CLASSIFICATION. JOURNAL OF SCIENCE AND APPLIED ENGINEERING, 9(1), 38-46. https://doi.org/10.31328/jsae.v9i1.7400

Vancouver

Agnesya PD, Rahajoe AD, Nurlaili AL. IMPLEMENTATION OF THE XGBOOST ALGORITHM FOR ACUTE RESPIRATORY INFECTION CLASSIFICATION. JOURNAL OF SCIENCE AND APPLIED ENGINEERING. 2026;9(1):38-46. doi:10.31328/jsae.v9i1.7400

BibTeX
@article{Agnesya2026,
  author = {Agnesya, Putri Dwi and Rahajoe, Ani Dijah and Nurlaili, Afina Lina},
  title = {IMPLEMENTATION OF THE XGBOOST ALGORITHM FOR ACUTE RESPIRATORY INFECTION CLASSIFICATION},
  journal = {JOURNAL OF SCIENCE AND APPLIED ENGINEERING},
  year = {2026},
  volume = {9},
  number = {1},
  pages = {38--46},
  doi = {10.31328/jsae.v9i1.7400}
}
Kostenlos starten →