{"id":"arxiv/2603.03107","url":"https://arxiv.org/abs/2603.03107","tags":["math.OC"],"score":0,"title":"Robust principal component analysis with rank and cardinality regularization under matrix factorization","domain":"math.OC","source":"arxiv","authors":["Wenjing Li","Wei Bian","Kim-Chuan Toh"],"reviews":0,"subject":"math.OC","summary":"Robust principal component analysis is an important representative method in data analysis. It is usually viewed as an optimization problem involving the rank and $\\ell_0$-norm of matrices. In this paper, we study the rank and $\\ell_0$ regularized optimization problem and its matrix factorization problem. We establish their equivalences on global minimizers and stationary points, respectively. Furthermore, we construct a broadly applicable equivalent nonconvex relaxation framework for the constrained factorization model in the sense of global minimizers and stationary points with strong…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-03-03","reactions":{}}