Show simple item record

dc.contributor.authorWang, Deqing
dc.date.accessioned2019-11-25T10:59:22Z
dc.date.available2019-11-25T10:59:22Z
dc.date.issued2019
dc.identifier.isbn978-951-39-7968-3
dc.identifier.urihttps://jyx.jyu.fi/handle/123456789/66505
dc.description.abstractElectroencephalography (EEG) is a powerful technique for the study of human brain and cognitive neuroscience. Nowadays, more and more EEG data are organized in high-dimension form, which is called tensor. Tensor decomposition is just the suitable tool to exploit the multiway data and extract EEG features that are linked to cognitive processes. Since the high-dimension EEG tensor often contains a large amount of data points, highly efficient tensor decomposition algorithm is desired. In addition, EEG tensor are sometimes nonnegative and the intrinsic features usually have some special properties, such as sparse. In order to extract meaningful feature components, it is necessary to incorporate constraint and regularization to tensor decomposition algorithm. In this dissertation, we study the CANDECOMP/PARAFAC (CP) tensor decomposition with both nonnegative constraint and sparse regularization, which is abbreviated as sparse NCP. An inexact block coordinate descent (BCD) framework is employed for the non-convex sparse NCP problem. Five optimization methods are employed to solve the sparse NCP, including multiplicative update (MU), alternating nonnegative least squares/quadratic programming (ANLS/ANQP), hierarchical altering least squares (HALS), alternating proximal gradient (APG) and alternating direction method of multipliers (ADMM), all of which are carefully tailored to the sparse regularization problem. In order to improve the stability, we also utilize proximal algorithm particularly for ANLS/ANQP and HALS. Applications on real-world EEG datasets are carried out. First, we use NCP to decompose a fifth-order event-related potential (ERP) tensor, which was collected by proprioceptive stimuli on human hands. Next, ongoing EEG tensors are analyzed using sparse NCP. The data were collected by naturalistic and continuous music stimulus. Finally, we analyze two modalities of ongoing EEG tensor and music signals simultaneously by N-way partial least square (N-PLS). In conclusion, our designed tensor decomposition methods with constraint and regularization are able to decompose high-order tensor data efficiently and extract meaningful EEG features linked to cognitive processes. Keywords: Tensor decomposition, nonnegative CANDECOMP/PARAFAC, sparse regularization, block coordinate descent, EEG data analysisen
dc.format.mimetypeapplication/pdf
dc.language.isoeng
dc.publisherJyväskylän yliopisto
dc.relation.ispartofseriesJYU dissertations
dc.relation.haspart<b>Artikkeli I:</b> Deqing Wang, Xiulin Wang, Tapani Ristaniemi and Fengyu Cong. Sparse Nonnegative Tensor Decomposition in Inexact Block Coordinate Descent Framework. <i>Submitted to a journal.</i>
dc.relation.haspart<b>Artikkeli II:</b> Wang, D., Cong, F., & Ristaniemi, T. (2019). Higher-order Nonnegative CANDECOMP/PARAFAC Tensor Decomposition Using Proximal Algorithm. In <i>ICASSP 2019 : Proceedings of the 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (pp. 3457-3461). IEEE.</i> <a href="https://doi.org/10.1109/ICASSP.2019.8683217"target="_blank"> DOI: 10.1109/ICASSP.2019.8683217</a>
dc.relation.haspart<b>Artikkeli III:</b> Wang, D., Zhu, Y., Ristaniemi, T., & Cong, F. (2018). Extracting multi-mode ERP features using fifth-order nonnegative tensor decomposition. <i>Journal of Neuroscience Methods, 308, 240-247.</i> <a href="https://doi.org/10.1016/j.jneumeth.2018.07.020"target="_blank"> DOI: 10.1016/j.jneumeth.2018.07.020</a>
dc.relation.haspart<b>Artikkeli IV:</b> Wang, D., Wang, X., Zhu, Y., Toiviainen, P., Huotilainen, M., Ristaniemi, T., & Cong, F. (2018). Increasing Stability of EEG Components Extraction Using Sparsity Regularized Tensor Decomposition. In <i>T. Huang, J. Lv, C. Sun, & A. V. Tuzikov (Eds.), ISNN 2018 : Advances in Neural Networks : 15th International Symposium on Neural Networks, Proceedings (pp. 789-799). Springer International Publishing.</i> <a href="https://doi.org/10.1007/978-3-319-92537-0_89"target="_blank"> DOI: 10.1007/978-3-319-92537-0_89</a>
dc.relation.haspart<b>Artikkeli V:</b> Wang, D., Cong, F., Zhao, Q., Toiviainen, P., Nandi, A. K., Huotilainen, M., . . . , & Cichocki, A. (2016). Exploiting ongoing EEG with multilinear partial least squares during free-listening to music. In <i>F. A. N. Palmieri, A. Uncini, K. Diamantaras, & J. Larsen (Eds.), Proceedings of MLSP 2016 : IEEE International Workshop on Machine Learning for Signal Processing. Institute of Electrical and Electronic Engineers.</i> <a href="https://doi.org/10.1109/MLSP.2016.7738849"target="_blank"> DOI: 10.1109/MLSP.2016.7738849</a>
dc.rightsIn Copyright
dc.subjectEEG
dc.subjectsignaalinkäsittely
dc.subjectsignaalianalyysi
dc.subjectalgoritmit
dc.subjectkognitiivinen neurotiede
dc.subjectaivotutkimus
dc.subjecttensor decomposition
dc.subjectnonnegative CANDECOMP/PARAFAC
dc.subjectsparse regularization
dc.subjectblock coordinate descent
dc.subjectEEG data analysis
dc.titleExtracting Meaningful EEG Features Using Constrained Tensor Decomposition
dc.typeDiss.
dc.identifier.urnURN:ISBN:978-951-39-7968-3
dc.contributor.yliopistoUniversity of Jyväskyläen
dc.contributor.yliopistoJyväskylän yliopistofi
dc.relation.issn2489-9003
dc.rights.copyright© The Author & University of Jyväskylä
dc.rights.accesslevelopenAccess
dc.type.publicationdoctoralThesis
dc.format.contentfulltext
dc.rights.urlhttps://rightsstatements.org/page/InC/1.0/


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record

In Copyright
Except where otherwise noted, this item's license is described as In Copyright