Parametric Dimensionality Reduction by Unsupervised Regression

Abstract

We introduce a parametric version (pDRUR) of the re-cently proposed Dimensionality Reduction by Unsupervised Regression algorithm. pDRUR alternately minimizes recon-struction error by fitting parametric functions given latent coordinates and data, and by updating latent coordinates given functions (with a Gauss-Newton method decoupled over coordinates). Both the fit and the update become much faster while attaining results of similar quality, and afford dealing with far larger datasets (105 points). We show in a number of benchmarks how the algorithm efficiently learns good latent coordinates and bidirectional mappings between the data and latent space, even with very noisy or low-quality initializations, often drastically improving the result of spectral and other methods.

Cite

Text

Carreira-Perpiñán and Lu. "Parametric Dimensionality Reduction by Unsupervised Regression." IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2010. doi:10.1109/CVPR.2010.5539862

Markdown

[Carreira-Perpiñán and Lu. "Parametric Dimensionality Reduction by Unsupervised Regression." IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2010.](https://mlanthology.org/cvpr/2010/carreiraperpinan2010cvpr-parametric/) doi:10.1109/CVPR.2010.5539862

BibTeX

@inproceedings{carreiraperpinan2010cvpr-parametric,
  title     = {{Parametric Dimensionality Reduction by Unsupervised Regression}},
  author    = {Carreira-Perpiñán, Miguel Á. and Lu, Zhengdong},
  booktitle = {IEEE/CVF Conference on Computer Vision and Pattern Recognition},
  year      = {2010},
  pages     = {1895-1902},
  doi       = {10.1109/CVPR.2010.5539862},
  url       = {https://mlanthology.org/cvpr/2010/carreiraperpinan2010cvpr-parametric/}
}