A1 Journal article (refereed)
Assessing the effectiveness of spatial PCA on SVM-based decoding of EEG data (2024)

Zhang, G., Carrasco, C. D., Winsler, K., Bahle, B., Cong, F., & Luck, S. J. (2024). Assessing the effectiveness of spatial PCA on SVM-based decoding of EEG data. Neuroimage, 293, Article 120625. https://doi.org/10.1016/j.neuroimage.2024.120625

JYU authors or editors

Publication details

All authors or editors: Zhang, Guanghui; Carrasco, Carlos D.; Winsler, Kurt; Bahle, Brett; Cong, Fengyu; Luck, Steven J.

Journal or series: Neuroimage

ISSN: 1053-8119

eISSN: 1095-9572

Publication year: 2024

Publication date: 03/05/2024

Volume: 293

Article number: 120625

Publisher: Elsevier

Publication country: United States

Publication language: English

DOI: https://doi.org/10.1016/j.neuroimage.2024.120625

Research data link: https://doi.org/10.18115/D5JW4R

Publication open access: Openly available

Publication channel open access: Open Access channel

Publication is parallel published (JYX): https://jyx.jyu.fi/handle/123456789/94848

Additional information: Additional data and scripts are available at https://osf.io/tgzew/

Abstract

Principal component analysis (PCA) has been widely employed for dimensionality reduction prior to multivariate pattern classification (decoding) in EEG research. The goal of the present study was to provide an evaluation of the effectiveness of PCA on decoding accuracy (using support vector machines) across a broad range of experimental paradigms. We evaluated several different PCA variations, including group-based and subject-based component decomposition and the application of Varimax rotation or no rotation. We also varied the numbers of PCs that were retained for the decoding analysis. We evaluated the resulting decoding accuracy for seven common event-related potential components (N170, mismatch negativity, N2pc, P3b, N400, lateralized readiness potential, and error-related negativity). We also examined more challenging decoding tasks, including decoding of face identity, facial expression, stimulus location, and stimulus orientation. The datasets also varied in the number and density of electrode sites. Our findings indicated that none of the PCA approaches consistently improved decoding performance related to no PCA, and the application of PCA frequently reduced decoding performance. Researchers should therefore be cautious about using PCA prior to decoding EEG data from similar experimental paradigms, populations, and recording setups.

Keywords: EEG; signal analysis; signal processing; principal component analysis; machine learning

Free keywords: EEG; MVPA; group-based PCA; subject-based PCA; dimensionality reduction

Contributing organizations

Ministry reporting: Yes

VIRTA submission year: 2024

Preliminary JUFO rating: 3