Number of the records: 1
Bayesian non-negative matrix factorization with adaptive sparsity and smoothness prior
- 1.
SYSNO ASEP 0500888 Document Type J - Journal Article R&D Document Type Journal Article Subsidiary J Článek ve WOS Title Bayesian non-negative matrix factorization with adaptive sparsity and smoothness prior Author(s) Tichý, Ondřej (UTIA-B) RID, ORCID
Bódiová, Lenka (UTIA-B)
Šmídl, Václav (UTIA-B) RID, ORCIDNumber of authors 3 Source Title IEEE Signal Processing Letters. - : Institute of Electrical and Electronics Engineers - ISSN 1070-9908
Roč. 26, č. 3 (2019), s. 510-514Number of pages 5 s. Publication form Print - P Language eng - English Country US - United States Keywords Non-negative matrix factorization ; Covariance matrix model ; Blind source separation ; Variational Bayes method ; Dynamic renal scintigraphy Subject RIV BB - Applied Statistics, Operational Research OECD category Automation and control systems R&D Projects GA18-07247S GA ČR - Czech Science Foundation (CSF) Method of publishing Limited access Institutional support UTIA-B - RVO:67985556 UT WOS 000458852100008 EID SCOPUS 85061747380 DOI 10.1109/LSP.2019.2897230 Annotation Non-negative matrix factorization (NMF) is generally an ill-posed problem which requires further regularization. Regularization of NMF using the assumption of sparsity is common as well as regularization using smoothness. In many applications it is natural to assume that both of these assumptions hold together. To avoid ad hoc combination of these assumptions using weighting coefficient, we formulate the problem using a probabilistic model and estimate it in a Bayesian way. Specifically, we use the fact that the assumptions of sparsity and smoothness are different forms of prior covariance matrix modeling. We use a generalized model that includes both sparsity and smoothness as special cases and estimate all its parameters using the variational Bayes method. The resulting matrix factorization algorithm is compared with state-of-the-art algorithms on large clinical dataset of 196 image sequences from dynamic renal scintigraphy. The proposed algorithm outperforms other algorithms in statistical evaluation. Workplace Institute of Information Theory and Automation Contact Markéta Votavová, votavova@utia.cas.cz, Tel.: 266 052 201. Year of Publishing 2020 Electronic address https://ieeexplore.ieee.org/document/8633424
Number of the records: 1