Классификация вариабельности сердечного ритма с использованием метода опорных векторов

Abstract

The time series of ECG signals, the analysis of normal and abnormal signals characteristics (the presence of tachycardia and arterial fibrillation) were investigated in this work. Proposed method based on machine learning which allows to classify these signals. The proposed method includes pre-processing of ECG signals, selection of signs and classification by using the method of reference vectors.