Heart Disease Prediction Using ANN Based Ensemble Framework (ANNEF) Approach

Abstract

The heart is the most important organ in the human body. One of the deadliest illnesses, heart disease is to blame for countless deaths worldwide. Because of the formation of fatty plaques in arteries, the heart fails to supply enough blood to other body areas to carry out its usual functions in heart disease (atherosclerosis). Age, a sedentary lifestyle, a family history of heart disease, stress, and other factors are all risk factors for heart disease. The World Health Organization estimates that 1.79 billion people die every year. If cardiac disease is detected early, the death rate can be reduced. In the diagnosis of cardiovascular disorders, machine learning plays an important role (CVDs). With the use of a Heart Disease Dataset, experimentation is undertaken using machine learning approaches to pick a better Heart Disease Prediction System (HDPS) (HDD). The HDD is made from publicly available datasets from the Cleveland, Hungary, Switzerland, VA Long Beach, and Statlog heart disease repositories. The final dataset contains 1190 records with 13 attributes such as age, gender, type of chest pain, resting blood pressure, serum cholesterol, fasting blood sugar, resting ECG, maximum heart rate, exerciseinduced angina, ST depression in patients' ECG, peak slope in ST-segment of ECG, number of vessels colored by fluoroscopy, and thalassemia to characterize heart diseases. The 14th attribute, 'Disease', on the other hand, is a Boolean variable that indicates whether the patient has heart disease or not, and it is chosen as a target attribute. Experimentation using ensemble frameworks referred to as ANNbased Ensemble Frameworks was done to improve the effectiveness of the heart disease prediction system (ANNEF). EKNN, K-Nearest Neighbors algorithm (KNN), Support Vector Machine (SVM), CART (Decision Tree method), and Logistic regression classifiers were used to create these ensemble frameworks. Instead of majority voting, this meta-output model is used to predict heart disease. The accuracy of ANNEF was enhanced to 95.69%.