Multi-view Metric Learning in Vector-valued Kernel Spaces

Riikka Huusari, Hachem Kadri, Cécile Capponi
Proceedings of the Twenty-First International Conference on Artificial Intelligence and Statistics, PMLR 84:415-424, 2018.

Abstract

We consider the problem of metric learning for multi-view data and present a novel method for learning within-view as well as between-view metrics in vector-valued kernel spaces, as a way to capture multi-modal structure of the data. We formulate two convex optimization problems to jointly learn the metric and the classifier or regressor in kernel feature spaces. An iterative three-step multi-view metric learning algorithm is derived from the optimization problems. In order to scale the computation to large training sets, a block-wise Nyström approximation of the multi-view kernel matrix is introduced. We justify our approach theoretically and experimentally, and show its performance on real-world datasets against relevant state-of-the-art methods.

Cite this Paper


BibTeX
@InProceedings{pmlr-v84-huusari18a, title = {Multi-view Metric Learning in Vector-valued Kernel Spaces}, author = {Riikka Huusari and Hachem Kadri and Cécile Capponi}, booktitle = {Proceedings of the Twenty-First International Conference on Artificial Intelligence and Statistics}, pages = {415--424}, year = {2018}, editor = {Amos Storkey and Fernando Perez-Cruz}, volume = {84}, series = {Proceedings of Machine Learning Research}, month = {09--11 Apr}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v84/huusari18a/huusari18a.pdf}, url = { http://proceedings.mlr.press/v84/huusari18a.html }, abstract = {We consider the problem of metric learning for multi-view data and present a novel method for learning within-view as well as between-view metrics in vector-valued kernel spaces, as a way to capture multi-modal structure of the data. We formulate two convex optimization problems to jointly learn the metric and the classifier or regressor in kernel feature spaces. An iterative three-step multi-view metric learning algorithm is derived from the optimization problems. In order to scale the computation to large training sets, a block-wise Nyström approximation of the multi-view kernel matrix is introduced. We justify our approach theoretically and experimentally, and show its performance on real-world datasets against relevant state-of-the-art methods.} }
Endnote
%0 Conference Paper %T Multi-view Metric Learning in Vector-valued Kernel Spaces %A Riikka Huusari %A Hachem Kadri %A Cécile Capponi %B Proceedings of the Twenty-First International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2018 %E Amos Storkey %E Fernando Perez-Cruz %F pmlr-v84-huusari18a %I PMLR %P 415--424 %U http://proceedings.mlr.press/v84/huusari18a.html %V 84 %X We consider the problem of metric learning for multi-view data and present a novel method for learning within-view as well as between-view metrics in vector-valued kernel spaces, as a way to capture multi-modal structure of the data. We formulate two convex optimization problems to jointly learn the metric and the classifier or regressor in kernel feature spaces. An iterative three-step multi-view metric learning algorithm is derived from the optimization problems. In order to scale the computation to large training sets, a block-wise Nyström approximation of the multi-view kernel matrix is introduced. We justify our approach theoretically and experimentally, and show its performance on real-world datasets against relevant state-of-the-art methods.
APA
Huusari, R., Kadri, H. & Capponi, C.. (2018). Multi-view Metric Learning in Vector-valued Kernel Spaces. Proceedings of the Twenty-First International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 84:415-424 Available from http://proceedings.mlr.press/v84/huusari18a.html .

Related Material