Multi-View Independent Component Analysis with Shared and Individual Sources
This work addresses the challenge of disentangling shared and individual latent sources in multi-view data, with applications in fields like bioinformatics, though it is incremental as it extends existing ICA frameworks to a specific noisy multi-view setting.
The paper tackles the problem of blind source separation in multi-view noisy linear ICA, where each view contains mixtures of shared and individual sources, by proving identifiability and developing a computational method to recover the sources. It demonstrates empirical recovery under noise, proposes a model selection procedure for shared source count, and applies it to transcriptome datasets to uncover shared sources for graph structure representation.
Independent component analysis (ICA) is a blind source separation method for linear disentanglement of independent latent sources from observed data. We investigate the special setting of noisy linear ICA where the observations are split among different views, each receiving a mixture of shared and individual sources. We prove that the corresponding linear structure is identifiable, and the source distribution can be recovered. To computationally estimate the sources, we optimize a constrained form of the joint log-likelihood of the observed data among all views. We also show empirically that our objective recovers the sources also in the case when the measurements are corrupted by noise. Furthermore, we propose a model selection procedure for recovering the number of shared sources which we verify empirically. Finally, we apply the proposed model in a challenging real-life application, where the estimated shared sources from two large transcriptome datasets (observed data) provided by two different labs (two different views) lead to recovering (shared) sources utilized for finding a plausible representation of the underlying graph structure.