Joint decompositions with flexible couplings

Rodrigo Cabral Farias (1) , Jérémy E Cohen (1) , Christian Jutten (2) , Pierre Comon (1)

1 GIPSA-CICS - GIPSA - Communication Information and Complex Systems
2 GIPSA-VIBS - GIPSA - Vision and Brain Signal Processing

A Bayesian framework is proposed to define flexible coupling models for joint decompositions of data sets. Under this framework, a solution to the joint decomposition can be cast in terms of a maximum a posteriori estimator. Examples of joint posterior distributions are provided , including general Gaussian priors and non Gaussian coupling priors. Then simulations are reported and show the effectiveness of this approach to fuse information from data sets, which are inherently of different size due to different time resolution of the measurement devices.

Keywords: multimodal data; data fusion; coupled decompositions; Tensor decompositions

Domains: Signal and Image Processing

Rodrigo Cabral Farias, Jérémy E Cohen, Christian Jutten, Pierre Comon. Joint decompositions with flexible couplings. LVA/ICA 2015 - 12th International Conference on Latent Variable Analysis and Signal Separation, Aug 2015, Liberec, Czech Republic. ⟨10.1007/978-3-319-22482-414⟩. ⟨hal-01135920v2⟩