Heart Disease Prediction Using Machine Learning Model Ensemble-Random Forest with Simple Regression

Abstract

Heterogeneous Criterion based Decision-making is the need of the hour when dealing with clinical data analysis such as Heart Disease data. A Machine Learning (ML) based model can effectively perform complex analysis and then predict the presence or absence of heart disease in an individual accurately. This analysis will aid the doctors in their diagnosis and save countless human lives. This paper proposes a hybrid model Random Forest with Simple Linear Regression (RFSLR) for classifying individuals as healthy individual or Heart Disease individual. The experimental analysis done by this work shows that RFSLR outperforms the existing classifiers Linear Regression(LR), Support Vector Machine (SVM), Random Forest (RF), Naïve Bayes (NB) and DTs(DT) in terms of performance metrics. This paper also recognizes the best features from the Cleveland Heart Disease Dataset (CHDD) using three Feature Selection Algorithm (FSA) RELIEF-F, LASSO and FOCUS.