An Ensemble Model With Genetic Algorithm for Classification of Coronary Artery Disease

Abstract

Coronary artery disease (CAD) is the most common form of heart disease and has become the primary reason for death. A correct and on-time diagnosis of CAD is very important. Diagnosis of CAD being a strenuous activity, scientists have planned different intelligent diagnostic frameworks for improved CAD diagnosis. Still, low CAD classification accuracy is an issue in these frameworks. In this paper, the authors propose a feature selection technique (FST) that utilizes a genetic algorithm (GA) with J48 classifier as the objective function to choose adequate features for better CAD classification accuracy. After feature removal, classification frameworks are used (i.e., artificial neural network [ANN]) like multilayer perceptron network (MLP), radial basis function network (RBFN), ANN-based ensemble model (ANN-EM), and deep neural network (DNN). Finally, this research proposes an integrated model of GA and ANN-EM for classification of CAD.