Structural Convergence Results for Approximation of Dominant Subspaces from Block Krylov Spaces.
Petros DrineasIlse C. F. IpsenEugenia-Maria KontopoulouMalik Magdon-IsmailPublished in: SIAM J. Matrix Anal. Appl. (2018)
Keyphrases
- structural information
- approximation algorithms
- conjugate gradient algorithm
- stationary points
- closed form
- structural features
- relative error
- linear computational complexity
- error bounds
- approximation spaces
- convergence rate
- convergence speed
- block size
- approximation schemes
- hilbert space
- approximation error
- data points
- high dimensional
- iterative algorithms
- faster convergence
- neural network
- parameter space
- approximation methods
- high dimensional data
- convergence analysis
- low dimensional
- block wise
- update rule
- principal component analysis