ECG Human Identification with Statistical Support Vector Machines

Electrocardiogram (ECG) as a biological information, it has some special feature. Different people will have different ECG information, even one person has different ECG when he is under different body state. In this paper we use the Electrocardiogram (ECG) to identify disease or to detect different person. Firstly, we collect the ECG information form different body state of the different people. Secondly we will preprocess the ECG data by using a method of statistical. Thirdly we can use the support vector machine to train the data, and then classify different people´s data into different class. And finally when there are one new ECG data, we can also use SVM to identify the new data. Because even one people have several ECG signal, with our statistical method, the classifier may gets better robust.