Innovative Approach: CNNs for Heart Disease Prediction

The fact that heart disease is still a health concern emphasizes the necessity of precise and effective techniques for early identification and diagnosis. The aim of this work is to investigate the general behavior of most machine learning algorithms. Analyzing a dataset comprising a spectrum of demographic data from people both with and without cardiac disease forms part of the study. Our suggested approach intends to assess and compare, for heart disease risk, the predictive capability of machine learning techniques. Among the methods under study are decision trees, random forests, support vector machines, logistic regression, k closest neighbors, and suggested CNN. The results suggest that various machine learning algorithms show levels of capabilities. Random Forests display accuracy and reliability while CNNs demonstrate their potential in capturing connections within the data and occasionally achieving higher accuracy than Random Forests. Support Vector Machines and Logistic Regression also exhibit performance. This study adds to the expanding research on predicting heart disease by providing an analysis of machine learning techniques.