On the Optimality of the Oja's Algorithm for Online PCA
arXiv:2104.00512 · doi:10.1007/s11222-023-10229-z
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.
arXiv comment25 pages. arXiv admin note: text overlap with arXiv:1711.06644
References in corpus (5)
- Minimax sparse principal subspace estimation in high dimensions
- Near-Optimal Stochastic Approximation for Online Principal Component Estimation
- Faster Eigenvector Computation via Shift-and-Invert Preconditioning
- An Improved Gap-Dependency Analysis of the Noisy Power Method
- Nearly Optimal Stochastic Approximation for Online Principal Subspace Estimation