Neural networks and the separation of cosmic microwave background and astrophysical signals in sky maps

We implement an independent component analysis (ICA) algorithm to separate signals of different origin in sky maps at several frequencies. Owing to its self-organizing capability, it works without pior assumptions on either the frequency dependence or the angualar power spectrum of the various signals; rather, it learns directly from the input data how to identify the statistically independent components, on the assumption that all but, at most, one of the components have non-Guassian distributions.