A Hybrid Machine Learning Technique Based Prediction of Blood Pressure

Blood pressure measurement is an essential tool in both identification of cardiovascular diseases and tracking diseases' progress, and improvements in ML methods have substantially improved blood pressure detection systems. In this paper we propose a novel method that utilizes features of both Random Forest algorithm and Naive Bayes algorithm in blood pressure level prediction. Because hypertension can cause cardiovascular diseases, precise prediction of blood pressure is essential to timely diagnose and treat the condition. The new hybrid model is designed using Random Forest and Naive Bayes in order to improve the level of predictive accuracy in the classification. Random Forest gives a strong numerical basis that is accurate, fast and optimal for Big Data; Naive Bayes makes probabilistic contribution and is also relatively simple. The development of hybrid model is based on numerous physiological and lifestyle data, which is used as training and testing set. This research therefore validates these hypotheses by showing that the hybrid approach yields better results as compared to the single models resulting in high levels of accuracy in blood pressure determination. This work explores the opportunity that salient feature real hybrid machine learning methodologies have for healthcare, with a specific focus on hypertension case, suggesting it as a lucrative tool for enhancing the condition's management and clinical consequences for patients. Simulation is conducted by running the python Spyder 3.7 software. The results that have been obtained show random forest has prediction accuracy of 86%, naive bayes has 89% and the proposed hybrid method has 92%.