Matching Pursuit Kernel Fisher Discriminant Analysis

Tom Diethe, Zakria Hussain, David Hardoon, John Shawe-Taylor
Proceedings of the Twelth International Conference on Artificial Intelligence and Statistics, PMLR 5:121-128, 2009.

Abstract

We derive a novel sparse version of Kernel Fisher Discriminant Analysis (KFDA) using an approach based on Matching Pursuit (MP). We call this algorithm Matching Pursuit Kernel Fisher Discriminant Analysis (MPKFDA). We provide generalisation error bounds analogous to those constructed for the Robust Minimax algorithm together with a sample compression bounding technique. We present experimental results on real world datasets, which show that MPKFDA is competitive with the KFDA and the SVM on UCI datasets, and additional experiments that show that the MPKFDA on average outperforms KFDA and SVM in extremely high dimensional settings.

Cite this Paper

BibTeX
@InProceedings{pmlr-v5-diethe09a, title = {Matching Pursuit Kernel Fisher Discriminant Analysis}, author = {Diethe, Tom and Hussain, Zakria and Hardoon, David and Shawe-Taylor, John}, booktitle = {Proceedings of the Twelth International Conference on Artificial Intelligence and Statistics}, pages = {121--128}, year = {2009}, editor = {van Dyk, David and Welling, Max}, volume = {5}, series = {Proceedings of Machine Learning Research}, address = {Hilton Clearwater Beach Resort, Clearwater Beach, Florida USA}, month = {16--18 Apr}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v5/diethe09a/diethe09a.pdf}, url = {https://proceedings.mlr.press/v5/diethe09a.html}, abstract = {We derive a novel sparse version of Kernel Fisher Discriminant Analysis (KFDA) using an approach based on Matching Pursuit (MP). We call this algorithm Matching Pursuit Kernel Fisher Discriminant Analysis (MPKFDA). We provide generalisation error bounds analogous to those constructed for the Robust Minimax algorithm together with a sample compression bounding technique. We present experimental results on real world datasets, which show that MPKFDA is competitive with the KFDA and the SVM on UCI datasets, and additional experiments that show that the MPKFDA on average outperforms KFDA and SVM in extremely high dimensional settings.} }
Endnote
%0 Conference Paper %T Matching Pursuit Kernel Fisher Discriminant Analysis %A Tom Diethe %A Zakria Hussain %A David Hardoon %A John Shawe-Taylor %B Proceedings of the Twelth International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2009 %E David van Dyk %E Max Welling %F pmlr-v5-diethe09a %I PMLR %P 121--128 %U https://proceedings.mlr.press/v5/diethe09a.html %V 5 %X We derive a novel sparse version of Kernel Fisher Discriminant Analysis (KFDA) using an approach based on Matching Pursuit (MP). We call this algorithm Matching Pursuit Kernel Fisher Discriminant Analysis (MPKFDA). We provide generalisation error bounds analogous to those constructed for the Robust Minimax algorithm together with a sample compression bounding technique. We present experimental results on real world datasets, which show that MPKFDA is competitive with the KFDA and the SVM on UCI datasets, and additional experiments that show that the MPKFDA on average outperforms KFDA and SVM in extremely high dimensional settings.
RIS
TY - CPAPER TI - Matching Pursuit Kernel Fisher Discriminant Analysis AU - Tom Diethe AU - Zakria Hussain AU - David Hardoon AU - John Shawe-Taylor BT - Proceedings of the Twelth International Conference on Artificial Intelligence and Statistics DA - 2009/04/15 ED - David van Dyk ED - Max Welling ID - pmlr-v5-diethe09a PB - PMLR DP - Proceedings of Machine Learning Research VL - 5 SP - 121 EP - 128 L1 - http://proceedings.mlr.press/v5/diethe09a/diethe09a.pdf UR - https://proceedings.mlr.press/v5/diethe09a.html AB - We derive a novel sparse version of Kernel Fisher Discriminant Analysis (KFDA) using an approach based on Matching Pursuit (MP). We call this algorithm Matching Pursuit Kernel Fisher Discriminant Analysis (MPKFDA). We provide generalisation error bounds analogous to those constructed for the Robust Minimax algorithm together with a sample compression bounding technique. We present experimental results on real world datasets, which show that MPKFDA is competitive with the KFDA and the SVM on UCI datasets, and additional experiments that show that the MPKFDA on average outperforms KFDA and SVM in extremely high dimensional settings. ER -
APA
Diethe, T., Hussain, Z., Hardoon, D. & Shawe-Taylor, J.. (2009). Matching Pursuit Kernel Fisher Discriminant Analysis. Proceedings of the Twelth International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 5:121-128 Available from https://proceedings.mlr.press/v5/diethe09a.html.

Related Material