作者: Meng Huanhuan , Zhang Yue
DOI: 10.1109/CSE.2014.36
关键词:
摘要: This paper introduces an electrocardiogram beat classification method based on deep belief networks. includes two parts: feature extraction and classification. In the part, features are extracted from original signal: including by networks timing interval features. Several classifiers selected to classify beat, nonlinear support vector machine with Gaussian kernel achieves best accuracy, reaching 98.49. Compared other similar methods classification, our can improve recognition performance of some types beats.