Header logo is

Joint Kernel Maps

2005

Conference Paper

ei


We develop a methodology for solving high dimensional dependency estimation problems between pairs of data types, which is viable in the case where the output of interest has very high dimension, e.g., thousands of dimensions. This is achieved by mapping the objects into continuous or discrete spaces, using joint kernels. Known correlations between input and output can be defined by such kernels, some of which can maintain linearity in the outputs to provide simple (closed form) pre-images. We provide examples of such kernels and empirical results.

Author(s): Weston, J. and Schölkopf, B. and Bousquet, O.
Book Title: Proceedings of the 8th InternationalWork-Conference on Artificial Neural Networks
Journal: Proceedings of the 8th International Work-Conference on Artificial Neural Networks (Computational Intelligence and Bioinspired System)
Volume: LNCS 3512
Pages: 176-191
Year: 2005
Day: 0
Editors: J Cabestany and A Prieto and F Sandoval
Publisher: Springer

Department(s): Empirical Inference
Bibtex Type: Conference Paper (inproceedings)

DOI: 10.1007/11494669_23
Event Name: IWANN 2005
Event Place: Vilanova i la Geltrú, Barcelona, Spain

Address: Berlin Heidelberg, Germany
Organization: Max-Planck-Gesellschaft
School: Biologische Kybernetik

Links: PostScript

BibTex

@inproceedings{3472,
  title = {Joint Kernel Maps},
  author = {Weston, J. and Sch{\"o}lkopf, B. and Bousquet, O.},
  journal = {Proceedings of the 8th International Work-Conference on Artificial Neural Networks (Computational Intelligence and Bioinspired System)},
  booktitle = {Proceedings of the 8th InternationalWork-Conference on Artificial Neural Networks},
  volume = {LNCS 3512},
  pages = {176-191},
  editors = {J Cabestany and A Prieto and F Sandoval},
  publisher = {Springer},
  organization = {Max-Planck-Gesellschaft},
  school = {Biologische Kybernetik},
  address = {Berlin Heidelberg, Germany},
  year = {2005},
  doi = {10.1007/11494669_23}
}