Anytime Representation Learning

Zhixiang Xu, Matt Kusner, Gao Huang, Kilian Weinberger
Proceedings of the 30th International Conference on Machine Learning, PMLR 28(3):1076-1084, 2013.

Abstract

Evaluation cost during test-time is becoming increasingly important as many real-world applications need fast evaluation (e.g. web search engines, email spam filtering) or use expensive features (e.g. medical diagnosis). We introduce Anytime Feature Representations (AFR), a novel algorithm that explicitly addresses this trade-off in the data representation rather than in the classifier. This enables us to turn conventional classifiers, in particular Support Vector Machines, into test-time cost sensitive anytime classifiers - combining the advantages of anytime learning and large-margin classification.

Cite this Paper

BibTeX
@InProceedings{pmlr-v28-xu13b, title = {Anytime Representation Learning}, author = {Xu, Zhixiang and Kusner, Matt and Huang, Gao and Weinberger, Kilian}, booktitle = {Proceedings of the 30th International Conference on Machine Learning}, pages = {1076--1084}, year = {2013}, editor = {Dasgupta, Sanjoy and McAllester, David}, volume = {28}, number = {3}, series = {Proceedings of Machine Learning Research}, address = {Atlanta, Georgia, USA}, month = {17--19 Jun}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v28/xu13b.pdf}, url = {https://proceedings.mlr.press/v28/xu13b.html}, abstract = {Evaluation cost during test-time is becoming increasingly important as many real-world applications need fast evaluation (e.g. web search engines, email spam filtering) or use expensive features (e.g. medical diagnosis). We introduce Anytime Feature Representations (AFR), a novel algorithm that explicitly addresses this trade-off in the data representation rather than in the classifier. This enables us to turn conventional classifiers, in particular Support Vector Machines, into test-time cost sensitive anytime classifiers - combining the advantages of anytime learning and large-margin classification.} }
Endnote
%0 Conference Paper %T Anytime Representation Learning %A Zhixiang Xu %A Matt Kusner %A Gao Huang %A Kilian Weinberger %B Proceedings of the 30th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2013 %E Sanjoy Dasgupta %E David McAllester %F pmlr-v28-xu13b %I PMLR %P 1076--1084 %U https://proceedings.mlr.press/v28/xu13b.html %V 28 %N 3 %X Evaluation cost during test-time is becoming increasingly important as many real-world applications need fast evaluation (e.g. web search engines, email spam filtering) or use expensive features (e.g. medical diagnosis). We introduce Anytime Feature Representations (AFR), a novel algorithm that explicitly addresses this trade-off in the data representation rather than in the classifier. This enables us to turn conventional classifiers, in particular Support Vector Machines, into test-time cost sensitive anytime classifiers - combining the advantages of anytime learning and large-margin classification.
RIS
TY - CPAPER TI - Anytime Representation Learning AU - Zhixiang Xu AU - Matt Kusner AU - Gao Huang AU - Kilian Weinberger BT - Proceedings of the 30th International Conference on Machine Learning DA - 2013/05/26 ED - Sanjoy Dasgupta ED - David McAllester ID - pmlr-v28-xu13b PB - PMLR DP - Proceedings of Machine Learning Research VL - 28 IS - 3 SP - 1076 EP - 1084 L1 - http://proceedings.mlr.press/v28/xu13b.pdf UR - https://proceedings.mlr.press/v28/xu13b.html AB - Evaluation cost during test-time is becoming increasingly important as many real-world applications need fast evaluation (e.g. web search engines, email spam filtering) or use expensive features (e.g. medical diagnosis). We introduce Anytime Feature Representations (AFR), a novel algorithm that explicitly addresses this trade-off in the data representation rather than in the classifier. This enables us to turn conventional classifiers, in particular Support Vector Machines, into test-time cost sensitive anytime classifiers - combining the advantages of anytime learning and large-margin classification. ER -
APA
Xu, Z., Kusner, M., Huang, G. & Weinberger, K.. (2013). Anytime Representation Learning. Proceedings of the 30th International Conference on Machine Learning, in Proceedings of Machine Learning Research 28(3):1076-1084 Available from https://proceedings.mlr.press/v28/xu13b.html.

Related Material