Detection of Sleep Staging from EEG: A Comparison of Feature Dimensionality Reduction and Classifier Selection

Xiaotong Ding, Lei Yang, Zhongcai Liao, Yanwen Fang

2022

Abstract

Sleep is an important part of maintaining human health. With the high incidence of sleep disorders, sleep has attracted much attention. Sleep staging is an effective means to study sleep structure. This paper studies the effect of different feature dimensionality reduction algorithms on the accuracy of sleep analysis, including the influence of principal component analysis, factor analysis and autoencoders on common classifiers, such as random forest and support vector machine for automated sleep stage detection. The combination with the highest accuracy was used to verify the sleep EEG data obtained in our laboratory. The results show that, using autoencoders to reduce dimension can keep the performance of the model, while using principal component analysis and factor analysis can improve the accuracy of the model in most cases.

Paper Citation

in Harvard Style

Ding X., Yang L., Liao Z. and Fang Y. (2022). Detection of Sleep Staging from EEG: A Comparison of Feature Dimensionality Reduction and Classifier Selection. In Proceedings of the 4th International Conference on Biotechnology and Biomedicine - Volume 1: ICBB; ISBN 978-989-758-637-8, SciTePress, pages 266-269. DOI: 10.5220/0012019300003633

in Bibtex Style

@conference{icbb22,
author={Xiaotong Ding and Lei Yang and Zhongcai Liao and Yanwen Fang},
title={Detection of Sleep Staging from EEG: A Comparison of Feature Dimensionality Reduction and Classifier Selection},
booktitle={Proceedings of the 4th International Conference on Biotechnology and Biomedicine - Volume 1: ICBB},
year={2022},
pages={266-269},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0012019300003633},
isbn={978-989-758-637-8},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 4th International Conference on Biotechnology and Biomedicine - Volume 1: ICBB
TI - Detection of Sleep Staging from EEG: A Comparison of Feature Dimensionality Reduction and Classifier Selection
SN - 978-989-758-637-8
AU - Ding X.
AU - Yang L.
AU - Liao Z.
AU - Fang Y.
PY - 2022
SP - 266
EP - 269
DO - 10.5220/0012019300003633
PB - SciTePress