Institute Homepage
Institute Homepage DE Sign In

Back

Empirical Inference Conference Paper 2007

Supervised Feature Selection via Dependence Estimation

no image
Empirical Inference
Arthur Gretton

We introduce a framework for filtering features that employs the Hilbert-Schmidt Independence Criterion (HSIC) as a measure of dependence between the features and the labels. The key idea is that good features should maximise such dependence. Feature selection for various supervised learning problems (including classification and regression) is unified under this framework, and the solutions can be approximated using a backward-elimination algorithm. We demonstrate the usefulness of our method on both artificial and real world datasets.

Author(s): Song, L. and Smola, AJ. and Gretton, A. and Borgwardt, KM. and Bedo, J.
Journal: Proceedings of the 24th Annual International Conference on Machine Learning (ICML 2007)
Pages: 823-830
Year: 2007
Month: June
Day: 0
Editors: Ghahramani, Z.
Publisher: ACM Press
Bibtex Type: Conference Paper (inproceedings)
Address: New York, NY, USA
DOI: 10.1145/1273496.1273600
Event Name: Twenty-Fourth Annual International Conference on Machine Learning (ICML 2007)
Event Place: Corvallis, OR, USA
Digital: 0
Electronic Archiving: grant_archive
ISBN: 978-1-59593-793-3
Language: en
Organization: Max-Planck-Gesellschaft
School: Biologische Kybernetik
Links:

BibTex 

@inproceedings{4462,
  title = {Supervised Feature Selection via Dependence Estimation},
  journal = {Proceedings of the 24th Annual International Conference on Machine Learning (ICML 2007)},
  abstract = {We introduce a framework for filtering features that employs the Hilbert-Schmidt Independence Criterion (HSIC) as a measure of dependence between the features and the labels. The key idea is that good features should maximise such dependence. Feature selection for various supervised learning problems (including classification and regression) is unified under this framework, and the solutions can be approximated using a backward-elimination algorithm. We demonstrate the usefulness of our method on both artificial and real world datasets.},
  pages = {823-830},
  editors = {Ghahramani, Z. },
  publisher = {ACM Press},
  organization = {Max-Planck-Gesellschaft},
  school = {Biologische Kybernetik},
  address = {New York, NY, USA},
  month = jun,
  year = {2007},
  slug = {4462},
  author = {Song, L. and Smola, AJ. and Gretton, A. and Borgwardt, KM. and Bedo, J.},
  month_numeric = {6}
}