Regularized Low Rank Approximation of Weighted Data Sets

Manuscript The manuscript on Regularized Low Rank Approximation of Weighted Data Sets can be downloaded here. Supporting Data This page contain supporting data and resources related to the work on "Regularized Low Rank Approximation of Weighted Data". The PDF version of the manuscript cannot render the high resolution images. Those images are available here. CODES: Download the tarball here. Click. With in the tarball, you will find the MATLAB routines, test data (used in the paper), and MATLAB scripts demonstrating the usage of the functions. Starting Materials: The VIMOS R band image used for illustration. Click. The mask (binary weights) to seperate celestial objects from the background. Click. Estimated Backgrounds: The background obtained by fitting tensor products of Legendre polynomial, the degrees of polynomials sums up to 4. Click. The background obtained with low rank (rank 4) approximation. Click. The background obtained with low rank (rank 4) approximation, and regularization with second order finite difference of accuracy 2. Click. The background obtained with low rank (rank 4) approximation, and regularization with second order finite difference of accuracy 8. Click. Corrected Image: Corrected image, where the background obtained by fitting tensor products of Legendre polynomial, the degrees of polynomials sums up to 4. Click. Corrected image, where the background obtained with low rank (rank 4) approximation. Click. Corrected image, where the background obtained with low rank (rank 4) approximation, and regularization with second order finite difference of accuracy 2. Click. Corrected image, where the background obtained with low rank (rank 4) approximation, and regularization with second order finite difference of accuracy 8. Click.