Advanced Predictive Modeling for Hypertension Risk Based on Health Indicators and Machine Learning Technique

Introduction: The condition of hypertension significantly accelerates the incidence of cardiovascular diseases and demands timely and proper measurement of the avoidable risk. Traditional techniques in the measurement of the pressure in the arteries provide accurate figures but are incapable of forecasting the risk of the development of hypertension. Aim: The goal was to establish an Efficient Pelican Optimized Dynamic Random Forest (EPO-DRF) model from health markers to forecast the hypertension probability. Methods: Patient information was extracted from clinical history, such as clinical predictors and lifestyle predictors of hypertension. Preprocessing, such as normalization and cleaning, was carried out to ensure precision and consistency. The significant predictors, such as age, cholesterol, blood sugar, and BMI, were determined. Optimum pelican optimization was used to increase the predictive efficiency by identifying the most significant predictors and removing redundant predictors. Result: To forecast the hypertension probability, the EPO-DRF model also displayed excellent outcomes, such as the F1-score (86.2%), the accuracy (90.4%), the sensitivity (87.5%), and the precision (85.7%). Classification performance and the most significant feature selection also underwent optimization in the course of the optimization to increase the efficacy of the model. Conclusion: The novel methodology arrived at an effective and efficient way to attain hypertension screening at an early stage, in alignment with preventive care practices and minimizing hypertension complications. It also helped healthcare analytics by having a precise predictive model to project future hypertension detection, making timely intervention and enhancing outcomes among the patients.