Active Nearest Neighbors in Changing Environments

Christopher Berlind, Ruth Urner
Proceedings of the 32nd International Conference on Machine Learning, PMLR 37:1870-1879, 2015.

Abstract

While classic machine learning paradigms assume training and test data are generated from the same process, domain adaptation addresses the more realistic setting in which the learner has large quantities of labeled data from some source task but limited or no labeled data from the target task it is attempting to learn. In this work, we give the first formal analysis showing that using active learning for domain adaptation yields a way to address the statistical challenges inherent in this setting. We propose a novel nonparametric algorithm, ANDA, that combines an active nearest neighbor querying strategy with nearest neighbor prediction. We provide analyses of its querying behavior and of finite sample convergence rates of the resulting classifier under covariate shift. Our experiments show that ANDA successfully corrects for dataset bias in multi-class image categorization.

Cite this Paper

BibTeX
@InProceedings{pmlr-v37-berlind15, title = {Active Nearest Neighbors in Changing Environments}, author = {Berlind, Christopher and Urner, Ruth}, booktitle = {Proceedings of the 32nd International Conference on Machine Learning}, pages = {1870--1879}, year = {2015}, editor = {Bach, Francis and Blei, David}, volume = {37}, series = {Proceedings of Machine Learning Research}, address = {Lille, France}, month = {07--09 Jul}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v37/berlind15.pdf}, url = {https://proceedings.mlr.press/v37/berlind15.html}, abstract = {While classic machine learning paradigms assume training and test data are generated from the same process, domain adaptation addresses the more realistic setting in which the learner has large quantities of labeled data from some source task but limited or no labeled data from the target task it is attempting to learn. In this work, we give the first formal analysis showing that using active learning for domain adaptation yields a way to address the statistical challenges inherent in this setting. We propose a novel nonparametric algorithm, ANDA, that combines an active nearest neighbor querying strategy with nearest neighbor prediction. We provide analyses of its querying behavior and of finite sample convergence rates of the resulting classifier under covariate shift. Our experiments show that ANDA successfully corrects for dataset bias in multi-class image categorization.} }
Endnote
%0 Conference Paper %T Active Nearest Neighbors in Changing Environments %A Christopher Berlind %A Ruth Urner %B Proceedings of the 32nd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2015 %E Francis Bach %E David Blei %F pmlr-v37-berlind15 %I PMLR %P 1870--1879 %U https://proceedings.mlr.press/v37/berlind15.html %V 37 %X While classic machine learning paradigms assume training and test data are generated from the same process, domain adaptation addresses the more realistic setting in which the learner has large quantities of labeled data from some source task but limited or no labeled data from the target task it is attempting to learn. In this work, we give the first formal analysis showing that using active learning for domain adaptation yields a way to address the statistical challenges inherent in this setting. We propose a novel nonparametric algorithm, ANDA, that combines an active nearest neighbor querying strategy with nearest neighbor prediction. We provide analyses of its querying behavior and of finite sample convergence rates of the resulting classifier under covariate shift. Our experiments show that ANDA successfully corrects for dataset bias in multi-class image categorization.
RIS
TY - CPAPER TI - Active Nearest Neighbors in Changing Environments AU - Christopher Berlind AU - Ruth Urner BT - Proceedings of the 32nd International Conference on Machine Learning DA - 2015/06/01 ED - Francis Bach ED - David Blei ID - pmlr-v37-berlind15 PB - PMLR DP - Proceedings of Machine Learning Research VL - 37 SP - 1870 EP - 1879 L1 - http://proceedings.mlr.press/v37/berlind15.pdf UR - https://proceedings.mlr.press/v37/berlind15.html AB - While classic machine learning paradigms assume training and test data are generated from the same process, domain adaptation addresses the more realistic setting in which the learner has large quantities of labeled data from some source task but limited or no labeled data from the target task it is attempting to learn. In this work, we give the first formal analysis showing that using active learning for domain adaptation yields a way to address the statistical challenges inherent in this setting. We propose a novel nonparametric algorithm, ANDA, that combines an active nearest neighbor querying strategy with nearest neighbor prediction. We provide analyses of its querying behavior and of finite sample convergence rates of the resulting classifier under covariate shift. Our experiments show that ANDA successfully corrects for dataset bias in multi-class image categorization. ER -
APA
Berlind, C. & Urner, R.. (2015). Active Nearest Neighbors in Changing Environments. Proceedings of the 32nd International Conference on Machine Learning, in Proceedings of Machine Learning Research 37:1870-1879 Available from https://proceedings.mlr.press/v37/berlind15.html.

Related Material