Sleep staging prediction model based on XGBoost

The sleep process is a complex process with dynamic changes, and the characteristics of sleep EEG signals in different sleep stages are also different. The XGBoost algorithm is used to extract and train the EEG signal characteristics of different sleep periods. Based on the principle of no repeated sampling, k-fold cross-validation is used to optimize the model. Finally, the model is retrained on all training sets with the parameters with the best generalization performance of the model, and the final evaluation is made using the test set independent of the training set. The accuracy of the sleep staging prediction model based on XGBoost can reach 90.6%.