Heart Disease Prediction System Using BIRCH Clustering

Diseases of all types that affect the heart or blood vessels, which lead to heart attacks, strokes and other heart defects known as heart diseases; nearly 630000 deaths occur in America due to heart diseases each year and 11.7% of adults in America have been diagnosed with heart disease which indicates more than one of every ten was affected with heart disease. Early recognition of heart disease will reduce the severe impact of the disease on the person. Efficient and more accurate disease prediction has been an important topic in the prediction system using machine learning. This paper aims to develop an Improved Heart Disease Prediction Model (IHDPM) with a new approach using Unsupervised and Supervised Learning techniques. The results showed that the combination of BIRCH clustering and decision tree classifier can give more accuracy, i.e. 83.33%, and the proposed model, i.e. IHDPM is fast and accurate compared to the other existing model.