Heart Disease Prediction Based On Machine Learning

Abstract

An accurate model for DBSCAN (Outlier detection and removal). And implementing KNN by predicting the suitable k value. While SMOTE-ENN is used to balance the training dataset. Gradient boosting is a technique where new models are made and used to forecast the residuals or error, then the scores are added to find the presence or absence of disease. And implementing KNN by predicting the suitable k value. The model was built using few publicly accessible datasets, Statlog, heart failure clinical records datasets and Cleveland. These respective models output was compared to Each other respectively.