Next: Non-linear decorrelation algorithms
Up: Algorithms for ICA
Previous: Preprocessing of the data
The pioneering work in  was
inspired by neural
networks. Their algorithm was based on canceling the
non-linear cross-correlations, see Section 4.3.3.
The non-diagonal terms of the matrix
are updated according
where g1 and g2 are some odd non-linear functions, and the yiare computed at every iteration as
are set to zero. The yi then give, after
convergence, estimates of the independent components.
Unfortunately, the algorithm converges only under
rather severe restrictions (see ).