<![CDATA[Hypertension remains a persistent and widespread health problem in the adult population, yet many cases go undetected due to limited early symptoms and reliance on conventional clinical assessment. This study aims to develop and evaluate a hypertension risk prediction model in adult patients using the Random Forest algorithm. This study employed a quantitative approach with an exploratory–predictive study design based on electronic secondary data, with a descriptive–analytical framework utilizing data mining techniques. The study population comprised all adult patients registered at selected healthcare facilities, while the sample consisted of 120 adult patients selected by purposive sampling from the hypertension risk dataset on Kaggle. The instrument used was a structured electronic medical record table, including age, gender, body mass index (BMI), blood pressure, and relevant medical history. The data underwent preprocessing and encoding, then were analyzed using the Random Forest algorithm on the Python platform with the scikitlearn library. Model performance was evaluated using accuracy, precision, recall, and F1score metrics. The results showed that the Random Forest model provided an accuracy of 87.5%, precision of 91.7%, recall of 84.6%, and F1 score of 88.0%, indicating a strong hypertension risk classification capability. The study concluded that Random Forest can be utilized as a reliable decision support system for early detection of hypertension risk in adult populations, especially when integrated with electronic medical records.]]>
Copyrights © 2026
