Guarantees of Riemannian Optimization for Low Rank Matrix Recovery by Ke Wei, Jian-Feng Cai, Tony F. Chan, Shingyu Leung
We establish theoretical recovery guarantees of a family of Riemannian optimization algorithms for low rank matrix recovery, which is about recovering an $m\times n$ rank $r$ matrix from $p < mn$ number of linear measurements. The algorithms are first interpreted as the iterative hard thresholding algorithms with subspace projections. Then based on this connection, we prove that if the restricted isometry constant $R_{3r}$ of the sensing operator is less than $C_\kappa /\sqrt{r}$ where $C_\kappa$ depends on the condition number of the matrix, the Riemannian gradient descent method and a restarted variant of the Riemannian conjugate gradient method are guaranteed to converge to the measured rank $r$ matrix provided they are initialized by one step hard thresholding. Empirical evaluation shows that the algorithms are able to recover a low rank matrix from nearly the minimum number of measurements necessary.
No comments:
Post a Comment