<![CDATA[The increasing incidence of Dengue Hemorrhagic Fever (DHF), Tuberculosis (TB), and Diarrhea in a district area highlights the urgent need for a data-driven prediction system to support public health policy. This study develops a predictive model of case numbers at the sub-district level using the Decision Tree Regression algorithm within the CRISP-DM methodology. Secondary data from 2020-2023 were utilized, including disease case records (Health Office), demographic data (BPS), and environmental data (BMKG). The system was implemented as a web-based application built with PHP and Python/Flask, enabling dataset management, model retraining, and interactive visualization of predictions, complemented by risk classification and recommended interventions. Experimental results demonstrate high predictive accuracy, with R² values of 0.9130 for TB, 0.8805 for DHF, and 0.8228 for Diarrhea. Overall, the proposed system serves as an objective and measurable decision-support tool, assisting the District Health Office in formulating preventive policies more rapidly and effectively.]]>
Copyrights © 2025
