Feature selection in SUMER spatial spectra using wavelet decomposition and ICA
Luis Manuel Sarro, Á. Berihuete, Coryn A. L. Bailer‐Jones · AIP conference proceedings · 2008
Here we briefly describe the techniques taken from multivariate statistics and data mining, used for the characterization of the solar spatial spectra in the database of observations taken with the SUMER instrument onboard SOHO. In particular, we describe the combination of wavelet decomposition and Independent Component Analysis in the separation of the active and quiet components.