Distinctive Analysis of Risk Forecasting Models for Coronary Heart Disease using Random Forests

Abstract

Coronary Heart Disease (CHD) represents a leading reason for global morbidity, thus prompting the requirement for precise risk forecasting models for urgent intervention. This research evaluates multiple machine learning models, with particular attention to Random Forest, to evaluate their success in predicting the likelihood of coronary heart disease. With preprocessing, feature selection applied GA and PSO, as well as model evaluation, the Framingham Heart Study dataset was used. In achieving an accuracy of 94.1% with a precision of 94.9% and $\mathbf{F 1}$ score of $\mathbf{7 8. 4 4 \%}$, The Random Forest model outperformed Logistic Regression and Support Vector Machines. The model suitably detected central risk factors and emphasized the complicated interactions that affect the emergence of CHD. The research indicates that Random Forest is a reasonable, interpretable, and clinically useful tool that strengthens early diagnosis and personalized treatment planning for CHD, perhaps improving patient outcomes. Keywords-coronary heart disease (CHD), early identification, machine learning models, random forest, framingham heart study dataset.

Ähnliche Arbeiten