Hybrid Ensemble and Features Selection for Heart Disease Prediction Using Machine Learning

Abstract

Diseases are very harmful, both physically and mentally. Despite of any external injury, diseases can affect the parts of organism. Diseases are always characterized by specific symptoms and signs. Arteria Coronaria Disease id related to blood flow obstruction. It is one of the most common disease in humans. Heart diseases are not predictable or expected. They often occur suddenly.  Machine learning techniques can be applied for prediction of heart disease. In this paper we have used the dataset of UCI repository for heart disease prediction using different parameters. After preprocessing and features selection we have used different machine learning classifiers to train the model. We have performed the heart disease prediction using various trained machine learning models.  Techniques like Logistic regression, Decision tree, Support Vector Machine, K Nearest Neighbor, Naïve Bayes and Random Forest algorithms are used in the prediction of heart disease and hybrid of these algorithms provides 83.16 % accuracy.