A PCG classification method using xgboost

According to the characteristics of PCG, wavelet packet energy features and MFCC features are extracted. The heartbeat signals are classified by xgboost algorithm. The experimental results show that the classification accuracy is significantly higher than that of the single feature when all the features are used. When the training set is 90% and the test set is 10%, the classification results are the best. The accuracy, sensitivity and specificity are 80.31%, 80.01% and 80.61%, respectively.