Fast algorithms for robust principal component analysis with an upper bound on the rank

The robust principal component analysis (RPCA) decomposes a data matrix into low-rank part and sparse part. There are mainly two types of algorithms for RPCA. first type algorithm applies regularization terms on the singular values to obtain matrix. However, calculating can be very expensive large matrices. second replaces as multiplication small They faster than because no value decomposition (SVD) is required. rank required, an accurate estimation needed reasonable solution. In this paper, we propose that combine both types. Our proposed require upper bound SVD First, they cost matrices negligible. Second, more instead exact Furthermore, apply Gauss-Newton method increase speed our algorithms. Numerical experiments show better performance