Comparative Analysis of Supervised Learning Algorithms for Clinical Diagnosis Heart Attack Prediction

This paper involves comparative research of novel deep machine learning and machine learning networks for the prediction of heart disease symptoms. Heart disease is one of the leading causes of death worldwide, and its early discovery saves a person's life. The objective of the research was to create a method of predicting heart attacks from lifestyle and clinical data through machine learning. The Machine learning algorithms that will be employed for prediction are discovered by conducting experiments that test combination of ML algorithms such as Logistic Regression, Random Forest, Support Vector Machines, and Gradient Boosting in the study. Data sets are downloaded from public sources and are processed to completion with missing value management, feature scaling, and exploratory data analysis (EDA). Validation has been performed using accuracy, precision, recall, and F1-score. The Random Forest classifier performed with an accuracy of 89%. The paper also discusses constraints and how to follow up the work, such as using larger data sets and new deep learning algorithms.