Cardiovascular Disease Prediction using Machine Learning Ensemble Methods

Abstract

Recently the main cause of death occurring due to cardiovascular disease has happened even in the past. Early diagnosis of the disease can be assessed to reduce the high risk and ensure healthiness. Data mining techniques have been significantly used as it helps in zero or less intervention of humans and it is seen as the best technique as it gives precise result with the best accuracy. The study is conducted on ensemble methods and built a model using boosting and bagging classifiers. The objective of this work is to design and implement a heart disease prediction system using machine learning ensemble methods namely, Random Forest, Adaptive Boosting, Gradient Boosting, and XGBoost. The effective performance of the applied ensemble techniques is analyzed, and a mobile application is developed for the same. The proposed mobile application is built as a user interface that accepts data based on clinical attributes concerning heart disease. This mainly helps in the medical field such as laboratories that incorporate the developed model. The outcome of the proposed model predicts the probability of a person suffering from heart disease. The accuracy of the models is evaluated.