2017/04/30 by Jérémy E. Cohen, Nicolas Gillis · 1 citation
Computer Science · Engineering · Mathematics · #Algorithm #Artificial intelligence #Blind Source Separation Techniques #Canonical correlation #Computer science #Decomposition #Identifiability #Interpretability #K-SVD #Machine learning #Mathematics #Matrix decomposition #Pattern recognition (psychology) #Physics #Pure mathematics #Sparse and Compressive Sensing Techniques #Sparse approximation #Tensor (intrinsic definition) #Tensor decomposition #Tensor decomposition and applications #Tucker decomposition #stat.ML
paper · pdf · doi:10.1109/tsp.2017.2777393
published as IEEE Trans. on Signal Processing 66 (7), pp. 1876-1889, 2018
arxiv created 2017/11/08 · openalex publication_date 2017/11/24 · arxiv updated 2018/03/13 · openalex created_date 2025/10/10 · openalex updated_date 2026/08/05
To ensure interpretability of extracted sources in tensor decomposition, we introduce in this paper a dictionary-based tensor canonical polyadic decomposition, which enforces one factor to belong exactly to a known dictionary. A new formulation of sparse coding is proposed, which enables high-dimensional tensors dictionary-based canonical polyadic decomposition. The benefits of using a dictionary in tensor decomposition models are explored both in terms of parameter identifiability and estimation accuracy. Performances of the proposed algorithms are evaluated on the decomposition of simulated data and the unmixing of hyperspectral images.