On the Optimality of the Oja's Algorithm for Online PCA
Machine Learning
2024-03-06 v2 Optimization and Control
Abstract
In this paper we analyze the behavior of the Oja's algorithm for online/streaming principal component subspace estimation. It is proved that with high probability it performs an efficient, gap-free, global convergence rate to approximate an principal component subspace for any sub-Gaussian distribution. Moreover, it is the first time to show that the convergence rate, namely the upper bound of the approximation, exactly matches the lower bound of an approximation obtained by the offline/classical PCA up to a constant factor.
Cite
@article{arxiv.2104.00512,
title = {On the Optimality of the Oja's Algorithm for Online PCA},
author = {Xin Liang},
journal= {arXiv preprint arXiv:2104.00512},
year = {2024}
}
Comments
25 pages. arXiv admin note: text overlap with arXiv:1711.06644