| Summary: | Introduction: Many different methods exist for identifying heart disease. This paper discusses developing a heart disease prediction application using machine learning algorithms to predict heart disease risk. The application aims to provide predictions to users, helping them assess their heart disease risk and make informed decisions regarding their health. Methods: The heart disease prediction application leverages the “Heart Disease UCI” dataset from Kaggle. The data are pre-processed, transformed, and split into 70% training and 30% testing sets. Prediction models are developed using three machine learning algorithms i.e. Support Vector Machine (SVM), Naïve Bayes, and k-nearest Neighbour (k-NN). Results: k-NN achieved an accuracy rate of 81.82%, Naive Bayes achieved 83.44%, and SVM achieved the highest accuracy rate of 84.74%. The results show that SVM outperformed the other algorithms. An application is then developed to implement the SVM prediction model. The application features various user interfaces, including sign-up and login pages for user registration and authentication. Users can enter their medical information, and the application uses the trained SVM model to predict their risk of heart disease. The results are presented to the users as a percentage risk and accompanied by appropriate health recommendations. Conclusion: The application may assist users in assessing heart disease risk and provide advice to minimize heart disease risk. © 2024 Universiti Putra Malaysia Press. All rights reserved.
|
|---|