Arabica: Robust ICA in a Pipeline

Authors:
Jarkko Ylipaavalniemi;Jyri Soppela
Affiliations:
Adaptive Informatics Research Centre, Department of Information and Computer Science, Helsinki University of Technology, Finland FI-02015 TKK;Adaptive Informatics Research Centre, Department of Information and Computer Science, Helsinki University of Technology, Finland FI-02015 TKK
Venue:
ICA '09 Proceedings of the 8th International Conference on Independent Component Analysis and Signal Separation
Year:
2009

Citing 1
Cited 3

Subspaces of spatially varying independent components in fMRI

ICA'07 Proceedings of the 7th international conference on Independent component analysis and signal separation

Denoising Single Trial Event Related Magnetoencephalographic Recordings

ICA '09 Proceedings of the 8th International Conference on Independent Component Analysis and Signal Separation
Consistency and asymptotic normality of FastICA and bootstrap FastICA

Signal Processing
Distributional convergence of subspace estimates in FastICA: a bootstrap study

LVA/ICA'12 Proceedings of the 10th international conference on Latent Variable Analysis and Signal Separation

Quantified Score

Hi-index	0.00

Visualization

Abstract

We introduce a new robust independent component analysis (ICA) toolbox for neuroinformatics, called "Arabica". The toolbox is designed to be modular and extendable also to other types of data. The robust ICA is the result of extensive research on reliable application of ICA to real-world measurements. The approach itself is based on sampling component estimates from multiple runs of ICA using bootstrapping. The toolbox is fully integrated to a recently developed processing pipeline environment, capable of running on a single machine or in a cluster of servers. Additionally, the toolbox works as a standalone package in Matlab, when the full pipeline is not required. The toolbox is aimed at being useful for both machine learning and neuroinformatics researchers.