Empirical Inference Conference Paper 2004

Semi-supervised kernel regression using whitened function classes

The use of non-orthonormal basis functions in ridge regression leads to an often undesired non-isotropic prior in function space. In this study, we investigate an alternative regularization technique that results in an implicit whitening of the basis functions by penalizing directions in function space with a large prior variance. The regularization term is computed from unlabelled input data that characterizes the input distribution. Tests on two datasets using polynomial basis functions showed an improved average performance compared to standard ridge regression.

Author(s): Franz, MO. and Kwon, Y. and Rasmussen, CE. and Schölkopf, B.
Book Title: Pattern Recognition, Proceedings of the 26th DAGM Symposium, Lecture Notes in Computer Science, Vol. 3175
Journal: Pattern Recognition, Proceedings of the 26th DAGM Symposium
Volume: LNCS 3175
Pages: 18-26
Year: 2004
Day: 0
Editors: CE Rasmussen and HH B{\"u}lthoff and MA Giese and B Sch{\"o}lkopf
Publisher: Springer
Bibtex Type: Conference Paper (inproceedings)
Address: Berlin, Gerrmany
Event Name: 26th DAGM Symposium
Event Place: Tübingen, Germany
Digital: 0
Electronic Archiving: grant_archive
Organization: Max-Planck-Gesellschaft
School: Biologische Kybernetik
Links:

BibTex

@inproceedings{2638,
  title = {Semi-supervised kernel regression using whitened function classes},
  journal = {Pattern Recognition, Proceedings of the 26th DAGM Symposium},
  booktitle = {Pattern Recognition, Proceedings of the 26th DAGM Symposium, Lecture Notes in Computer Science, Vol. 3175},
  abstract = {The use of non-orthonormal basis functions in ridge regression leads to an often undesired non-isotropic prior in function space. In this study, we investigate an alternative regularization technique that results in an implicit whitening of the basis functions by penalizing
  directions in function space with a large prior variance. The regularization term is computed from unlabelled input data that characterizes the input distribution. Tests on two datasets using polynomial basis functions showed an improved average performance compared to standard ridge regression.},
  volume = {LNCS 3175},
  pages = {18-26},
  editors = {CE Rasmussen and HH B{\"u}lthoff and MA Giese and B Sch{\"o}lkopf},
  publisher = {Springer},
  organization = {Max-Planck-Gesellschaft},
  school = {Biologische Kybernetik},
  address = {Berlin, Gerrmany},
  year = {2004},
  slug = {2638},
  author = {Franz, MO. and Kwon, Y. and Rasmussen, CE. and Sch{\"o}lkopf, B.}
}