Shared Independent Component Analysis for Multi-Subject Neuroimaging
arXiv:2110.13502
Abstract
We consider shared response modeling, a multi-view learning problem where one wants to identify common components from multiple datasets or views. We introduce Shared Independent Component Analysis (ShICA) that models each view as a linear transform of shared independent components contaminated by additive Gaussian noise. We show that this model is identifiable if the components are either non-Gaussian or have enough diversity in noise variances. We then show that in some cases multi-set canonical correlation analysis can recover the correct unmixing matrices, but that even a small amount of sampling noise makes Multiset CCA fail. To solve this problem, we propose to use joint diagonalization after Multiset CCA, leading to a new approach called ShICA-J. We show via simulations that ShICA-J leads to improved results while being very fast to fit. While ShICA-J is based on second-order statistics, we further propose to leverage non-Gaussianity of the components using a maximum-likelihood method, ShICA-ML, that is both more accurate and more costly. Further, ShICA comes with a principled method for shared components estimation. Finally, we provide empirical evidence on fMRI and MEG datasets that ShICA yields more accurate estimation of the components than alternatives.
arXiv commentAccepted at NeurIPS 2021
References in corpus (7)
- Scikit-learn: Machine Learning in Python
- Faster independent component analysis by preconditioning with Hessian approximations
- Independent Vector Analysis: Identification Conditions and Performance Bounds
- A Convolutional Autoencoder for Multi-Subject fMRI Data Aggregation
- Beyond Pham's algorithm for joint diagonalization
- CanICA: Model-based extraction of reproducible group-level ICA patterns from fMRI time series
- mvlearn: Multiview Machine Learning in Python