6533b7d1fe1ef96bd125d602

RESEARCH PRODUCT

Group Nonnegative Matrix Factorization with Sparse Regularization in Multi-set Data

Tapani RistaniemiWenya LiuFengyu CongXiulin Wang

subject

Computer scienceGroup (mathematics)020206 networking & telecommunications02 engineering and technologySparse approximationNon-negative matrix factorizationSet (abstract data type)Constraint (information theory)Computer Science::Computer Vision and Pattern Recognition0202 electrical engineering electronic engineering information engineeringSource separation020201 artificial intelligence & image processingJoint (audio engineering)Sparse regularizationAlgorithm

description

Constrained joint analysis of data from multiple sources has received widespread attention for that it allows us to explore potential connections and extract meaningful hidden components. In this paper, we formulate a flexible joint source separation model termed as group nonnegative matrix factorization with sparse regularization (GNMF-SR), which aims to jointly analyze the partially coupled multi-set data. In the GNMF-SR model, common and individual patterns of particular underlying factors can be extracted simultaneously with imposing nonnegative constraint and sparse penalty. Alternating optimization and alternating direction method of multipliers (ADMM) are combined to solve the GNMF-SR model. Using the experiment of simulated fMRI-like data, we demonstrate the ADMM-based GNMF-SR algorithm can achieve the better performance.

yearjournalcountryeditionlanguage
2021-01-242020 28th European Signal Processing Conference (EUSIPCO)
https://doi.org/10.23919/eusipco47968.2020.9287756