Data-Driven Hypertension Prediction Using Machine Learning Algorithms

Abstract

Hypertension is one of the most prevalent chronic diseases worldwide and a major risk factor for cardiovascular disorders. Early detection of hypertension is essential for reducing complications and improving patient outcomes. However, traditional diagnostic approaches often fail to identify high-risk individuals at an early stage. This study proposes a machine learning-based framework for the prediction of hypertension using Random Forest (RF), Random Committee (RC), and Multilayer Perceptron (MLP) classifiers. A primary dataset was collected and preprocessed to train and evaluate the predictive models. The Random Committee algorithm combines multiple randomized classifiers through majority voting to improve prediction stability and accuracy. Model performance was assessed using standard evaluation metrics, including accuracy, precision, recall, F1-score, and ROC analysis. Experimental results revealed that ensemble learning methods provide superior predictive performance. Among the tested models, Random Committee achieved the highest accuracy of 93.68%, followed by Random Forest with 92.34% accuracy. The findings demonstrate that machine learning techniques can effectively support early hypertension prediction, enabling timely intervention, better clinical decision-making, and improved preventive healthcare management.