Multilayer Perceptron Networks For Aid In The Diagnosis Of Coronary Heart Disease

Résumé: Heart disease, the predominant cause of morbidity and mortality worldwide, poses considerable diagnostic challenges. The imperative to identify and assess risks at an early stage requires the development of a robust and effective forecasting system. Machine learning enables accurate predictions, proving invaluable in various fields such as finance and healthcare. However, the varied performance of algorithms introduces complexities, with some achieving high prediction accuracy and others demonstrating comparatively lower precision. In this work, we introduce a method using a multi-layer perceptron neural network to provide decision support in the diagnosis of heart diseases in patients. This proposed model is systematically compared with four important machine learning counterparts: Random Forest, Logistic Regression, Naive Bayes, and SVM. Performance evaluations are conducted using a comprehensive dataset derived from a cardiovascular study including Massachusetts residents with coronary heart disease. The obtained results suggest that the accuracy of the model based on a multilayer perceptron is superior compared to the four other machine learning models. The use of MLPs appears to be a promising and effective approach for enhancing diagnostic accuracy in the field of cardiovascular health when employing this type of data.

Publié dans la revue: Journal of Molecular and Pharmaceutical Sciences