PPG Signal Identification Method Based on CSASVM

Abstract

PPG signal is the inherent physiological signal of human body. It contains a lot of physiological and pathological information of human body. There are great differences in PPG signals among different individuals, and it has good uniqueness and confidentiality. This paper proposes an identity recognition method that uses the Crow Search Algorithm (CSA) to optimize the support vector machine (SVM) classification model. The method uses the matching pursuit (MP) sparse decomposition algorithm to sparse the individual PPG signals to extract the feature values of the PPG signals, and then uses the Relief algorithm to filter out the 16 main features to form feature vectors. The CSA algorithm is used to find the optimal parameters and create the optimal support vector machine classification model. The experiment proves that compared with the traditional SVM model, the accuracy of the optimization is increased by 10%-15%, and the recognition rate is 97.5%.