<![CDATA[Alternative medicine based on Traditional Chinese Medicine such as acupuncture, cupping, and herbs is widely used in stroke rehabilitation, but its effectiveness varies among patients. This study aims to compare the performance of three machine learning algorithms—Logistic Regression, Random Forest, and XGBoost—in predicting the success of alternative medicine for stroke patients. The dataset consists of 1040 medical records with 15 clinical features from a health center in Tangerang. Research methods include data preprocessing, 80:20 data splitting, model training, and evaluation using accuracy, precision, recall, F1-score, and Cohen’s Kappa metrics. Results show that Random Forest achieved the best performance with 93.91% accuracy, outperforming XGBoost (92.31%) and Logistic Regression (88.14%). Random Forest also demonstrated robustness with default configuration compared to XGBoost, which requires intensive hyperparameter tuning for optimal performance. These findings recommend Random Forest as a practical choice for clinical prediction system implementation with limited resources. The research contributes to the development of data-based decision support systems in stroke alternative medicine services]]>
Copyrights © 2026
