Automatic Arrhythmia Detection Using Wavelet Transform and RUSBoosted Trees Classification

It can be said that heart diseases are very common and fatal diseases. Therefore, it is necessary to determine the heart diseases correctly by examining the ECG In this study, it was aimed for detection of the arrhythmia by automatic examination of the person's electrocardiography (ECG) records. In this process, features obtained from the wavelet method was classified by RUSBoosted Trees method. The heartbeats were divided into seven different classes. RUSBoost method was used to reduce the number of features, and it speeds up the signal processing process. This method is known as bringing together a lot of weak learners and creating powerful learners from this process. ST-Petersburg Institute of Cardiological Database has been preferred for analysis. Test and training accuracy was found in 12 channel ECG data. The method was fast enough to detect real-time arrhythmia. MATLAB was used for all analyzes