ML Anthology

An Expectation-Maximization Approach to Nonlinear Component Analysis

Roman Rosipal, Mark A. Girolami
NeCo 2001 pp. 505-510
doi:10.1162/089976601300014439 /neco/2001/rosipal2001neco-expectationmaximization/

Abstract

The proposal of considering nonlinear principal component analysis as a kernel eigenvalue problem has provided an extremely powerful method of extracting nonlinear features for a number of classification and regression applications. Whereas the utilization of Mercer kernels makes the problem of computing principal components in, possibly, infinite-dimensional feature spaces tractable, there are still the attendant numerical problems of diagonalizing large matrices. In this contribution, we propose an expectation-maximization approach for performing kernel principal component analysis and show this to be a computationally efficient method, especially when the number of data points is large.

NeCo Semantic Scholar

Cite

Text

Rosipal and Girolami. "An Expectation-Maximization Approach to Nonlinear Component Analysis." Neural Computation, 2001. doi:10.1162/089976601300014439

Markdown

[Rosipal and Girolami. "An Expectation-Maximization Approach to Nonlinear Component Analysis." Neural Computation, 2001.](https://mlanthology.org/neco/2001/rosipal2001neco-expectationmaximization/) doi:10.1162/089976601300014439

BibTeX

@article{rosipal2001neco-expectationmaximization,
  title     = {{An Expectation-Maximization Approach to Nonlinear Component Analysis}},
  author    = {Rosipal, Roman and Girolami, Mark A.},
  journal   = {Neural Computation},
  year      = {2001},
  pages     = {505-510},
  doi       = {10.1162/089976601300014439},
  volume    = {13},
  url       = {https://mlanthology.org/neco/2001/rosipal2001neco-expectationmaximization/}
}