Nonnegative Tensor Train Factorization with DMRG Technique
- 8 Downloads
Tensor train is one of the modern decompositions used as low-rank tensor approximations of multidimensional arrays. If the original data is nonnegative we sometimes want the approximant to keep this property. In this work new methods for nonnegative tensor train factorization are proposed. Low-rank approximation approach helps to speed up the computations whereas DMRG technique allows to adapt nonnegative TT ranks for better accuracy. The performance analysis of the proposed algorithms as well as comparison with other nonnegative TT factorization method are presented.
Keywords and phrasesnonnegative tensor train multiplicative updates DMRG
Unable to display preview. Download preview PDF.
The author thanks E. E. Tyrtyshnikov for useful advices and participation in discussions of the results.
This article contains the results of the project performed in the framework of the implementation of the programs of the Central Competences of the National Technological Database “Center for Big Data Storage and Analysis” (project “Tensor methods for processing and analysis of Big Data”) of Lomonosov MSU with the Project Support Funding of the National Technological Reporting dated December 11, 2018, no. 13/1251/2018.
- 5.D. Lee and H. Seung, “Algorithms for non-negative matrix factorization,” Adv. Neural Inform. Process. Syst. 13 (2001).Google Scholar
- 14.TDALAB-MATLAB Toolbox for High-order Tensor Data Decompositions and Analysis. Current Vers. 1.1 (2013). https://github.com/andrewssobral/TDALAB.
- 16.F. Cong, A. Phan, Q. Zhao, T. Huttunen-Scott, J. Kaartinen, T. Ristaniemi, H. Lyytinen, and A. Cichocki, “Benefits of multi-domain feature of mismatch negativity extracted by nonnegative tensor factorization from low-density array EEG,” Int. J. Neural Syst. 22 (6) (2012).Google Scholar
- 17.E. Tyrtyshnikov and E. Shcherbakova, “Nonnegative tensor train factorizations and some applications,” Lect. Notes Comput. Sci. (in press).Google Scholar
- 18.D. Savostyanov and I. Oseledets, “Fast adaptive interpolation of multi-dimensional arrays in tensor train format,” in Proceedings of the 7th International Workshop on Multidimensional (nD) Systems, nDS 2011 (2011), pp. 1–8.Google Scholar
- 19.E. Tyrtyshnikov and E. Shcherbakova, “Nonnegative matrix factorization methods based on low-rank cross approximations,” Comput. Math. Math. Phys. 59 (8) (2019, in press).Google Scholar