CNN model cardiac beat classification method based on intra-class and inter-class distance

The invention relates to the technical field of ECG cardiac beat classification, in particular to a CNN model cardiac beat classification method based on intra-class and inter-class distance. The CNNmodel cardiac beat classification method based on intra-class and inter-class distance includes: constructing the heart beat characteristic mathematical model; based on the mathematical model of cardiac beat characteristics, constructing the common CNN model. Based on the mathematical model of cardiac beat characteristics and the generic CNN model, the generic CNN model is improved by using the method of minimizing the inter-class distance and maximizing the class spacing to construct the individualized CNN model. The cardiac beats are classified by synthesizing the output information of the personality CNN model. The invention establishes a common CNN model for extracting the common features between classes of the heart beat, and extracts the class individual features of the heart beat bythe individual CNN model