An experiment-based investigation into machine learning for predicting coronary heart disease

Abstract

Extensive inquiry has been conducted to explore potential applications of machine learning methodologies in the realm of cardiovascular disease management. To facilitate a more comprehensive investigation This study explores machine learning algorithms, specifically Support Vector Machines (SVM) and Artificial Neural Networks (ANN), for disease identification, focusing on cardiovascular diseases. Utilizing a Kaggle dataset of around seventy thousand medical records, the research aims to refine methodology and assess performance variations. SVM and ANN techniques are applied to the Kaggle dataset, revealing SVM accuracies of 0.9997 (default), 0.9998 (RBF kernel, C=100.0), and 1.000 (linear kernel, C=1000.0). The Feedforward neural network, using Adam optimization across 50 batches and 10 epochs, achieved perfect accuracy of 1.000.