<![CDATA[The prevalence of hypertension is currently increasing among adolescents. Despite numerous efforts to improve hypertension prevention, there are still limited approaches capable of accurately predicting hypertension prevention behaviors. Machine learning is needed to develop predictive models that can identify key predictors. This study aimed to develop and evaluate a machine learning model for predicting hypertension prevention behaviors among students in Islamic boarding schools. A cross-sectional design was employed in this study. Primary quantitative data were collected through validated questionnaires from 378 students, aged 15–18 years at three Islamic boarding schools in Jember, Indonesia. The data were analyzed using a machine learning approach involving data preprocessing, selection of indicator variables, and division of the dataset into training and testing datasets to develop and evaluate a predictive model of hypertension prevention behavior. The results showed that the machine learning–based predictive model of hypertension prevention behavior performed well, achieving an area under the curve (AUC) of 0.72, an accuracy of 95%, and a precision of 70%. The model identified competence, autonomy, and subjective norms as the main predictors and adequately distinguished between students with good and poor hypertension prevention behaviors. The machine learning approach performs better by providing a preprocessing phase, comprehensive model performance evaluation metrics, and new or previously unseen data to assess the model's generalizability. Future studies should extend the study to various settings and populations to improve generalizability. The predictive model can be used to predict hypertension prevention behavior using a number of independent variables.]]>
Copyrights © 2026
