<![CDATA[Tuberculosis (TB) remains a major public health challenge that requires continuous monitoring of patient medication adherence to ensure treatment success. However, conventional monitoring methods often rely on manual supervision, making it difficult for healthcare workers to track patient adherence effectively. This study aims to develop a mobile-based smart reminder system for monitoring tuberculosis medication adherence at Puskesmas Ariodillah and to classify patient adherence levels using the Support Vector Machine (SVM) algorithm. The experimental results show that the dataset consists of 503 records categorized as recovered patients and 397 records categorized as patients continuing medication. The SVM model achieved an average cross-validation accuracy of 69.03%. Furthermore, kernel comparison results indicate that the Radial Basis Function (RBF) kernel produces the lowest error rate compared to linear, polynomial, and sigmoid kernels, demonstrating better classification performance. The developed smart reminder system successfully supports tuberculosis medication monitoring and provides useful information regarding patient adherence. These findings indicate that integrating mobile health technology with machine learning can improve adherence monitoring and assist healthcare workers in identifying patients who require additional attention during treatment.]]>
Copyrights © 2026
