SleepEGAN: A GAN-enhanced Ensemble Deep Learning Model for Imbalanced Classification of Sleep Stages

TL;DR

SleepEGAN is a novel GAN-based ensemble deep learning approach that addresses class imbalance and heterogeneity in sleep EEG data, significantly improving sleep stage classification accuracy.

Contribution

The paper introduces a new GAN architecture for EEG data augmentation and a cost-free ensemble strategy to enhance sleep stage classification performance.

Findings

Improves classification accuracy over existing methods

Effective data augmentation for minority sleep stages

Enhanced robustness against data heterogeneity

Abstract

Deep neural networks have played an important role in automatic sleep stage classification because of their strong representation and in-model feature transformation abilities. However, class imbalance and individual heterogeneity which typically exist in raw EEG signals of sleep data can significantly affect the classification performance of any machine learning algorithms. To solve these two problems, this paper develops a generative adversarial network (GAN)-powered ensemble deep learning model, named SleepEGAN, for the imbalanced classification of sleep stages. To alleviate class imbalance, we propose a new GAN (called EGAN) architecture adapted to the features of EEG signals for data augmentation. The generated samples for the minority classes are used in the training process. In addition, we design a cost-free ensemble learning strategy to reduce the model estimation variance caused by the heterogeneity between the validation and test sets, so as to enhance the accuracy and robustness of prediction performance. We show that the proposed method can improve classification accuracy compared to several existing state-of-the-art methods using three public sleep datasets.