Heart disease prediction based on machine learning algorithms

Abstract

Heart disease is a medical research field in which the outcome can benefit lots of people. Because there are several factors that might raise the risk of heart disease, it is useful to build a prediction model to assist people in assessing their health. This paper makes use of a Kaggle dataset that was derived from CDC (Centers for Disease Control and Prevention). First, 8 components are analyzed using diagrams, and then the dataset is used to train classifiers in machine learning models. This paper conducts a comparative study between different algorithms, including Decision Tree, Logistic Regression, SVM (Support Vector Machine), and Random Forest. Besides, the factors taken into consideration while evaluating performance include accuracy, precision, recall, and f1-score. As a result, the maximum accuracy is reached by SVM with a linear kernel, and logistic regression achieves the highest precision. In addition, the highest recall and f1-score are obtained from the model SVM with an RBF kernel.