Nonlinear blind source separation using kernel feature spaces
Stefan Harmeling, Kawanabe M Ziehe A, K-R Müller, Lee, Tzyy‐Ping Jung, Scott Makeig, Terrence J. Sejnowski · 2001
In this work we propose a kernel-based blind source separation (BSS) algorithm that can perform nonlinear BSS for general invertible nonlinearities. For our kTDSEP algorithm we have to go through four steps: (i) adapting to the intrinsic dimension of the data mapped to feature space , (ii) finding an orthonormal basis of this submanifold, (iii) mapping the data into the subspace of spanned by this orthonormal basis, and (iv) applying temporal decorrelation BSS (TDSEP) to the mapped data. After demixing we get a number of irrelevant components and the original sources. To find out which ones are the components of interest, we propose a criterion that allows to identify the original sources. The excellent performance of kTDSEP is demonstrated in experiments on nonlinearly mixed speech data.