Convolutional Neural Network Monitoring of Sleep Characteristics of Senile Dementia Patients Using EEG Big Data Analysis

摘要

Aiming at the problem that the available sleep EEG data are all imbalanced small data sets and the staging effect of direct migration application of Convolutional Neural Network model is poor, a deep end-to-end automatic sleep staging model for a small amount of imbalanced original sleep EEG data sets is proposed from two aspects of data set reconstruction and model training optimization. The research results show that the model can realize end-to-end learning of a small amount of original sleep EEG data, and the overall classification effect is better than that of recent high-level models. Due to its end-to-end characteristics, compared with the traditional machine learning model based on feature engineering, the proposed model is more suitable for split personalized sleep monitoring equipment equipped with remote servers.