Computer Science Preprint Server (CSPS) http://www.sciencedirect.com/preprintarchive: Computational Intelligence/0304002 (2003-04-09) (bibtex)

CuBICA: Independent component analysis by simultaneous third- and fourth-order cumulant diagonalization.

Tobias Blaschke and Laurenz Wiskott


Abstract: CuBICA, an improved method for independent component analysis (ICA) based on the diagonalization of cumulant tensors is proposed. It is based on Comon's algorithm [Comon, 1994] but it takes third- and fourth-order cumulant tensors into account simultaneously. The underlying contrast function is also mathematically much simpler and has a more intuitive interpretation. It is therefore easier to optimize and approximate. A comparison with Comon's and three other ICA-algorithms on different data sets demonstrates its performance.


Relevant Project:


April 11, 2003, Laurenz Wiskott, http://www.neuroinformatik.ruhr-uni-bochum.de/PEOPLE/wiskott/