A Comparative Study using Random Forests and SVM Models for Machine Learning Techniques to Predict Heart Disease Accurately

Abstract

This paper analyzes the possible use of machine learning (ML) algorithms for early detection and prevention of cardiovascular disease (CVD) which is one of the leading global health concerns. Traditional methods of diagnosis face issues with the intricate and non-linear clinical data and forecasting the disease fails in the absence of accurate predictive tools. Hence, this work studies various ML algorithms like Support Vector Machines (SVM), Neural Networks, and hybrid forms like SVM– Random Forest and SVM–Neural Networks. Among these, the hybrid Random Forest and Neural Network model exhibited the clinical potential for the highest accurate predictive potential at 91.5%. This research illustration of hybrid ML model empowering advanced CVD predictive tools for early diagnosis and CVD management is built on 303 clinical records. After model implementation in Python, a set of model validation metrics which included precision and accuracy, were performed followed by testing. Overall, this research highlights the reliability of hybrid ML models for the advanced and efficient management of cardiovascular disease.

Ähnliche Arbeiten