fastHICA: Hierarchical Independent Component Analysis: a Multi-Scale Sparse Non-Orthogonal Data-Driven Basis
It implements HICA (Hierarchical Independent Component Analysis) algorithm. This approach, obtained through the integration between treelets and Independent Component Analysis, is able to provide a multi-scale non-orthogonal data-driven basis, whose elements have a phenomenological interpretation according to the problem under study.
- Piercesare Secchi, Simone Vantini, and Paolo Zanini
- Date of publication
- 2015-05-12 11:39:40
- Paolo Zanini <email@example.com>
- GPL (>= 2)
- Construction of the HICA basis
- Energy criterion
- Extraction of score and loading matrices.
- Estimate of the similarity matrix
Files in this package