Robust Low-Rank Matrix Completion via an Alternating Manifold Proximal Gradient Continuation Method

Explore a lecture on robust low-rank matrix completion using an alternating manifold proximal gradient continuation method. Delve into the formulation of this problem as a nonsmooth Riemannian optimization over Grassmann manifold, and learn about the proposed alternating manifold proximal gradient continuation method for solving it. Examine the convergence rate analysis and discover the advantages of this approach through numerical results on synthetic data and real-world applications in background extraction from surveillance videos. Gain insights into topics such as robust PCA, low-rank factorization, proximal gradient methods, and manifold optimization techniques.

Introduction
Robust PCA
LowRank Factorization
RTRMC Model
Proximal Gradient Method
Manifold Proximal Gradient
Convergence Results
mpg Algorithm
Variance
Numerical Results
Recovery Rate
Background Exception
Absolute Matrix
Conclusion
Questions