Smartphone-Based Heart Disease Classification Using Machine Learning Techniques

Abstract

Patients having heart diseases are diagnosed with a severe delay at times and further diagnosis in the absence of medical personnel can be fatal if the prediction is inaccurate. Therefore, this paper proposes the use of heart disease datasets to predict heart disease using various machine learning methods (Logistic Regression, Naive Bayes, Random Forest, k-nearest Neighbor, Support Vector Machine, Decision Tree Classifier, XGBoost Classifier, Artificial Neural Network). Cleveland, Hungarian, Switzerland, Long Beach VA and Statlog (Heart) datasets were used in this study which has 11 features of 1190 instances. The dataset was split into train and test sets with a ratio of 80:20. The performance was evaluated based on the accuracy, precision, recall, and F1 score for each of the models. From the eight models, the XGBoost Classifier outperformed other models with an accuracy of 93.7%. The trained model was integrated with the Android Studio framework to create the mobile application for the classification of heart disease.