An application of the Independent Component Analysis methodology to gamma ray astrophysical imaging

Abstract

Independent Component Analysis (ICA) is a statistical method often used to decompose a complex dataset in its independent sub-parts. It is a powerful technique to solve a typical Blind Source Separation problem. A fast calculation of the gamma ray sky observed by GLAST, assuming the expected instrumental response, has been implemented. The simulated images were used to test the capability of the ICA method in identifying the sources.