Classification of Imbalanced Cardiac Arrhythmia Data

Arrhythmias are irregularities in the heartbeat and can be life-threatening. Early diagnosis of Cardiac Arrhythmia is quite crucial for saving patient lives. In this study, main goal to detect presence cardiac arrhythmia classify it into 16 groups from ECG recordings. The dataset UCI databank used apply different network structures classification. number sample each class not same dataset. has a very immoderate distribution, moreover, some classes don't exist. imbalance condition between causes decrement performance classifier such as low classification accuracy. Also, cross-validation steps, data divided which includes samples overcome difficulty five satisfy that condition. training test datasets obtained combination these groups. To deal with dataset, first, typical algorithms Multilayer Perceptron (MLP), Support Vector Machine (SVM), Radial Basis Function (RBF), Random Forest (RF) data. According precision accuracy measurements classifiers class, nested constructed improve overall tried obtain better performance. results classical proposed four new ensemble networks presented compare their result shows random forest best terms and, even having highest almost results. For reason, planned increase enhancement future work.