ML Anthology

Stability of Oja's PCA Subspace Rule

Juha Karhunen
NeCo 1994 pp. 739-747
doi:10.1162/NECO.1994.6.4.739 /neco/1994/karhunen1994neco-stability/

Abstract

This paper deals with stability of Oja's symmetric algorithm for estimating the principal component subspace of the input data. Exact conditions are derived for the gain parameter on which the discrete algorithm remains bounded. The result is extended for a nonlinear version of Oja's algorithm.

NeCo Semantic Scholar

Cite

Text

Karhunen. "Stability of Oja's PCA Subspace Rule." Neural Computation, 1994. doi:10.1162/NECO.1994.6.4.739

Markdown

[Karhunen. "Stability of Oja's PCA Subspace Rule." Neural Computation, 1994.](https://mlanthology.org/neco/1994/karhunen1994neco-stability/) doi:10.1162/NECO.1994.6.4.739

BibTeX

@article{karhunen1994neco-stability,
  title     = {{Stability of Oja's PCA Subspace Rule}},
  author    = {Karhunen, Juha},
  journal   = {Neural Computation},
  year      = {1994},
  pages     = {739-747},
  doi       = {10.1162/NECO.1994.6.4.739},
  volume    = {6},
  url       = {https://mlanthology.org/neco/1994/karhunen1994neco-stability/}
}