0443717 20240111140902.120150526235959.9 84946099230 000427402902056 10.1109/ICASSP.2015.7178355 5 s. C cav_un_epca*0444627978-1-4673-6997-81520-61492169-2173PiscatawayIEEE Computer Society2015 tensor decomposition CANDECOMP/PARAFAC tensor deconvolution cav_un_auth*0274170 Phan A. H. JP cav_un_auth*0101212 Tichavský Petr UTIA-B Stochastická informatika Department of Stochastic Informatics SI SI Department of Stochastic Informatics Ústav teorie informace a automatizace AV ČR, v. v. i. cav_un_auth*0272321 Cichocki A. JP http://library.utia.cas.cz/separaty/2015/SI/tichavsky-0443717.pdf 277 kB cav_un_auth*0303443 GA14-13713S GA ČR CZ In this paper, we propose a low-rank tensor deconvolution problem which seeks multiway replicative patterns and corresponding activating tensors of rank-1. An alternating least squares (ALS) algorithm has been derived for the model to sequentially update loading components and the patterns. In addition, together with a good initialisation method using tensor diagonalization, the update rules have been implemented with a low cost using fast inversion of block Toeplitz matrices as well as an efficient update strategy. Experiments show that the proposed model and the algorithm are promising in feature extraction and clustering. cav_un_auth*0316577 2015 IEEE International Conference on Acoustics, Speech, and Signal Processing ICASSP 2015 19.04.2015-24.04.2015 Brisbane AU BB 10000 10100 10103 2016 3 PO RVO:67985556 http://hdl.handle.net/11104/0246782 cav_un_auth*0303002 RIKEN JP S RIV/67985556:_____/15:00443717!RIV16-AV0-67985556 191684107 oprava UT WOS RIV/67985556:_____/15:00443717!RIV16-GA0-67985556 191719243 oprava UT WOS n.a. Proceedings Paper Acoustics|Engineering Electrical Electronic 2015 84946099230 SCOPUS 000427402902056 WOS 277 kB cav_un_epca*0444627 Proceedings of 2015 IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP) 978-1-4673-6997-8 1520-6149 2169 2173 Piscataway IEEE Computer Society 2015