模糊最小二乘支持向量机在心电图分类中的应用研究  

Application of fuzzy least squares support vector machine in ECG classification

作　　者：刘卫华[1] 

机构地区：[1]甘肃政法学院公安技术学院,甘肃兰州730070

出　　处：《自动化与仪器仪表》2015年第12期162-164,共3页Automation & Instrumentation

摘　　要：心电图(Electrocardiogram,ECG)是诊断心血管疾病的重要依据。本文提出一种基于模糊最小二乘支持向量机(FLSSVM)的ECG分类诊断方法,并利用麻省理工学院(MIT-BIH)的心电图数据库中的数据进行训练和测试,通过在每个训练样本点中加入模糊隶属度,训练得到分类模型,仿真结果表明,FLSSVM分类器相比标准LSSVM和SVM,在分类正确率,分类速度以及适用的样本规模上都表现出了十足的优越性,FLSSVM分类器能够有效地处理实际分类问题。Electrocardiogram is an important approach to diagnose cardiovascular disease. The ECG classification diagnosis method based on Fuzzy Least Squares Support Vector Machines (FLSSVM) is proposed. The method, by adding fuzzy membership in each training sample point, obtains classification model by training. The experiment results show that FLSSVM classifier is better than LSSVM classifier and SVM classifier in speed and accuracy of classification when training and testing the data in MIT-BIH ar-rhythmia database.The FLSSVM classifier can deal with the actual classification problem effectively.

关 键 词：心电图 分类问题 模糊最小二乘支持向量机 模糊隶属度 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程]