Title :
Non-Negative Matrix Factorizations of Spontaneous Electroencephalographic Signals for Classification
Author :
Mingyu, Liu ; Jue, Wang ; Chongxun, Zheng
Author_Institution :
Key Lab. of Biomed. Inf. Eng. of Minist. of Educ., Xi´´an Jiaotong Univ.
Abstract :
Non-negative matrix factorization (NMF) is an algorithm that is able to learn a parts-based representation. The paper proposes a new spontaneous EEG classification method for attention-related tasks. NMF was employed as feature extraction tool, which leads to more localized and sparse features than other two reference methods: power spectrum method and principal component analysis. With conventional back propagation neural network classifier, several experiments were carried out. It was showed that the NMF-ANN structure preserved the spatio-temporal characteristics of EEG signals
Keywords :
backpropagation; electroencephalography; feature extraction; matrix decomposition; medical signal processing; neural nets; signal classification; signal representation; spatiotemporal phenomena; ANN; EEG classification; attention-related tasks; backpropagation neural network classifier; feature extraction; nonnegative matrix factorizations; parts-based representation; power spectrum method; principal component analysis; signal classification; spatiotemporal characteristics; spontaneous electroencephalographic signals; Artificial neural networks; Biomedical measurements; Covariance matrix; Data mining; Electroencephalography; Feature extraction; Karhunen-Loeve transforms; Matrix decomposition; Principal component analysis; Vectors; Attention Deficit Hyperactivity Disorder; Back Propagation Artificial Neural Network; Electroencephalogram; Non-Negative Matrix Factorization; Principal component analysis;
Conference_Titel :
Engineering in Medicine and Biology Society, 2005. IEEE-EMBS 2005. 27th Annual International Conference of the
Conference_Location :
Shanghai
Print_ISBN :
0-7803-8741-4
DOI :
10.1109/IEMBS.2005.1617052
