A Perturbation Bound on the Subspace Estimator from Canonical Projections
Machine Learning
2022-06-30 v1 Information Theory
Machine Learning
math.IT
Abstract
This paper derives a perturbation bound on the optimal subspace estimator obtained from a subset of its canonical projections contaminated by noise. This fundamental result has important implications in matrix completion, subspace clustering, and related problems.
Keywords
Cite
@article{arxiv.2206.14278,
title = {A Perturbation Bound on the Subspace Estimator from Canonical Projections},
author = {Karan Srivastava and Daniel L. Pimentel-Alarcón},
journal= {arXiv preprint arXiv:2206.14278},
year = {2022}
}
Comments
To appear in Proc. of IEEE, ISIT 2022