Heart disease diagnosis using extreme learning based neural networks

Abstract

Heart disease is the leading cause of death in Indonesia based on 2010 Hospital Information System (SIRS) Report. Early detection and treatment of heart disease will reduce the patient mortality rate. Therefore, implementation of artificial neural networks (ANN) technique in diagnosing heart disease have been widely used and reached good accuracy. Beside of that, there are disadvantages in implementation of ANN technique, such as a long training process, many parameters have to be tuned, the obtained solution potentially get stuck in local minima, and activation function must be differentiable. We implemented Extreme Learning Machine (ELM) which is fast, simple tuning, and better generalization model learning algorithm. It has better performance than backpropagation ANN, Support Vector Machine (SVM), and decision tree. The results indicate that the ELM model has potentially implemented to help medical professional in diagnosing heart disease.