We consider the problem of multi-task learning, that is, learning multiple related functions. Our approach is based on a hierarchical Bayesian framework, that exploits the equivalence between parametric linear models and nonparametric Gaussian processes (GPs). The resulting models can be learned easily via an EM-algorithm. Empirical studies on multi-label text categorization suggest that the presented models allow accurate solutions of these multi-task problems.
%0 Conference Paper
%1 yu_learning_2005
%A Yu, Kai
%A Tresp, Volker
%A Schwaighofer, Anton
%B Proceedings of the 22nd international conference on Machine learning
%C Bonn, Germany
%D 2005
%I ACM
%K
%P 1012-1019
%R 10.1145/1102351.1102479
%T Learning Gaussian processes from multiple tasks
%U http://portal.acm.org/citation.cfm?id=1102479&dl=GUIDE&coll=GUIDE&CFID=69344422&CFTOKEN=94303924
%X We consider the problem of multi-task learning, that is, learning multiple related functions. Our approach is based on a hierarchical Bayesian framework, that exploits the equivalence between parametric linear models and nonparametric Gaussian processes (GPs). The resulting models can be learned easily via an EM-algorithm. Empirical studies on multi-label text categorization suggest that the presented models allow accurate solutions of these multi-task problems.
%@ 1-59593-180-5
@inproceedings{yu_learning_2005,
abstract = {We consider the problem of multi-task learning, that is, learning multiple related functions. Our approach is based on a hierarchical Bayesian framework, that exploits the equivalence between parametric linear models and nonparametric Gaussian processes (GPs). The resulting models can be learned easily via an EM-algorithm. Empirical studies on multi-label text categorization suggest that the presented models allow accurate solutions of these multi-task problems.},
added-at = {2009-03-05T08:50:36.000+0100},
address = {Bonn, Germany},
author = {Yu, Kai and Tresp, Volker and Schwaighofer, Anton},
biburl = {https://www.bibsonomy.org/bibtex/2db96238c10869835cab8ad79f44b8d45/bcao},
booktitle = {Proceedings of the 22nd international conference on Machine learning},
doi = {10.1145/1102351.1102479},
interhash = {a3dcc2f9502736d2349eca4ccc556357},
intrahash = {db96238c10869835cab8ad79f44b8d45},
isbn = {1-59593-180-5},
keywords = {},
pages = {1012-1019},
publisher = {ACM},
timestamp = {2009-03-05T08:50:36.000+0100},
title = {Learning Gaussian processes from multiple tasks},
url = {http://portal.acm.org/citation.cfm?id=1102479\&dl=GUIDE\&coll=GUIDE\&CFID=69344422\&CFTOKEN=94303924},
year = 2005
}