Classification of ECG Arrhythmia using Artificial Intelligence techniques (RBF and SVM)

Electrocardiogram (ECG) is a test that measures the electrical performance of the heart. It is one of the most important tests in the field of medicine because we used it to detect all heart problems. To observe the results of this test we applied Artificial Neural Network exactly Radial-Based Functional Network (RBF) and Support Vector Machine (SVM) which belongs to supervised machine-learning approaches. These algorithms are used to predict the classification of ECG signals. Each of these techniques is exploited for classification and regression purposes. The goal is to classify the normal and abnormal beats in ECG signals with a very low error rate and good accuracy. A comparison between the value of the accuracy and roc curves acquired during this work gives us an idea about the effectiveness of each approach in the classification of heartbeats. In this work MIT-BIH Arrhythmia database is exploited.