摘要
在对心电信号进行离散小波变换并提取优化特征组合的基础上,利用标准算法(l-a-r算法)和二叉树算法分别构建支持向量机分类器实现心电图的分类,对不同小波下提取不同维特征向量构建的分类器性能进行比较,同时对取自MIT-BIH数据库的4类心电图(正常心搏、左束支传导阻滞心搏、右束支传导阻滞心搏和起搏心搏)进行分类.结果表明,采用标准算法对db2小波下8维特征向量训练的支持向量机分类器分类性能最优,总体分类正确率达98.770/0.
Wavelet transform is applied to electrocardiograph （ECG） beat and the optimized feature combinations are obtained by feature searching algorithm. Support vector machine（SVM） classifiers are trained by using l-against-rest （1-a-r） algorithm and binary tree algorithm for ECG beat classification. The capabilities of classitiers using different feature vectors with different wavelets are compared. Four types of ECG beats （normal beat, left bundle branch block beat, right bundle branch block beat and paced beat） obtained from MIT-BIH database are classified by the algorithms. The results show that the classifier trained by 8 dimensional feature vectors based on db2 wavelet using 1-a-r algorithm has the best performance with an accuracy of 98.77%.
出处
《天津大学学报》
EI
CAS
CSCD
北大核心
2007年第7期811-815,共5页
Journal of Tianjin University(Science and Technology)
基金
天津市自然科学基金资助项目(06YFSYSF02200)
关键词
心搏分类
小波变换
特征提取
支持向量机
eleetroeardiograph（ECG） beat classification
wavelet transform
feature extraction
support vector machine
