Natural Gradient Algorithm for Blind Source Separation with Unknown Source Number

Shao Lianlia · Electronic Science and Technology · 2015

A summary of the natural gradient algorithm for blind source separation with an unknown number of sources is presented and the conclusion is drawn that it is the redundant movement of separating matrix in the zero space of the transpose of mixing matrix that causes the divergence. By the projected natural gradient algorithm,we theoretically prove that with the increase in the number of iterations,the norm of the redundant subjects to an exponential distribution.

Read the paper · More papers on PaperTik