Deep Gaussian Mixture-Hidden Markov Model for Classification of EEG Signals

Electroencephalography (EEG) signals are complex dynamic phenomena that exhibit nonlinear and nonstationary behaviors. These characteristics tend to undermine the reliability of existing hand-crafted EEG features that ignore time-varying information and impair the performances of classification mode...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:IEEE transactions on emerging topics in computational intelligence Ročník 2; číslo 4; s. 278 - 287
Hlavní autoři: Wang, Min, Abdelfattah, Sherif, Moustafa, Nour, Hu, Jiankun
Médium: Journal Article
Jazyk:angličtina
Vydáno: Piscataway IEEE 01.08.2018
The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Témata:
autoencoder
Autoregressive models
Brain modeling
Classification
deep learning
EEG classification
Electroencephalography
Feature extraction
Gaussian mixture model
hidden Markov model
Hidden Markov models
Indexing
Machine learning
Markov analysis
Markov chains
Markov processes
Signal classification
Task analysis
time-series
ISSN:2471-285X, 2471-285X
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:Electroencephalography (EEG) signals are complex dynamic phenomena that exhibit nonlinear and nonstationary behaviors. These characteristics tend to undermine the reliability of existing hand-crafted EEG features that ignore time-varying information and impair the performances of classification models. In this paper, we propose a novel method that can automatically capture the nonstationary dynamics of EEG signals for diverse classification tasks. It consists of two components. The first component uses an autoregressive-deep variational autoencoder model for automatic feature extraction, and the second component uses a Gaussian mixture-hidden Markov model for EEG classification with the extracted features. We compare the performance of our proposed method and the state-of-the-art methods in two EEG classification tasks, subject, and event classification. Results show that our approach outperforms the others by averages of <inline-formula><tex-math notation="LaTeX">\text{15}\%\pm \text{6.3}</tex-math> </inline-formula> (p-value <inline-formula><tex-math notation="LaTeX"><\text{0.05}</tex-math></inline-formula>) and <inline-formula><tex-math notation="LaTeX">\text{22}\%\pm \text{5.7}</tex-math></inline-formula> (p-value <inline-formula><tex-math notation="LaTeX"><\text{0.05}</tex-math></inline-formula>) for subject and event classifications, respectively.
Bibliografie:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:2471-285X
2471-285X
DOI:10.1109/TETCI.2018.2829981