Updating TULV decomposition
Abstract
A truncated ULV decomposition (TULVD) of an m x n matrix A of rank k is a decomposition of the form A = U1LV1T + E, where U-1 and V-1 are left orthogonal matrices, L is a lower triangular matrix and E is an error matrix. We present an updating algorithm of order O (nk) that reveals the rank correctly and produces good approximation to the subspaces of the matrix A.