Baseline Methods for Active Learning

Gavin C. Cawley
; Active Learning and Experimental Design workshop In conjunction with AISTATS 2010, JMLR Workshop and Conference Proceedings 16:47-57, 2011.

Abstract

In many potential applications of machine learning, unlabelled data are abundantly available at low cost, but there is a paucity of labelled data, and labeling unlabelled examples is expensive and/or time-consuming. This motivates the development of active learning methods, that seek to direct the collection of labelled examples such that the greatest performance gains can be achieved using the smallest quantity of labelled data. In this paper, we describe some simple pool-based active learning strategies, based on optimally regularised linear [kernel] ridge regression, providing a set of baseline submissions for the Active Learning Challenge. A simple random strategy, where unlabelled patterns are submitted to the oracle purely at random, is found to be surprisingly effective, being competitive with more complex approaches.

Cite this Paper


BibTeX
@InProceedings{pmlr-v16-cawley11a, title = {Baseline Methods for Active Learning}, author = {Gavin C. Cawley}, booktitle = {Active Learning and Experimental Design workshop In conjunction with AISTATS 2010}, pages = {47--57}, year = {2011}, editor = {Isabelle Guyon and Gavin Cawley and Gideon Dror and Vincent Lemaire and Alexander Statnikov}, volume = {16}, series = {Proceedings of Machine Learning Research}, address = {Sardinia, Italy}, month = {16 May}, publisher = {JMLR Workshop and Conference Proceedings}, pdf = {http://proceedings.mlr.press/v16/cawley11a/cawley11a.pdf}, url = {http://proceedings.mlr.press/v16/cawley11a.html}, abstract = {In many potential applications of machine learning, unlabelled data are abundantly available at low cost, but there is a paucity of labelled data, and labeling unlabelled examples is expensive and/or time-consuming. This motivates the development of active learning methods, that seek to direct the collection of labelled examples such that the greatest performance gains can be achieved using the smallest quantity of labelled data. In this paper, we describe some simple pool-based active learning strategies, based on optimally regularised linear [kernel] ridge regression, providing a set of baseline submissions for the Active Learning Challenge. A simple random strategy, where unlabelled patterns are submitted to the oracle purely at random, is found to be surprisingly effective, being competitive with more complex approaches.} }
Endnote
%0 Conference Paper %T Baseline Methods for Active Learning %A Gavin C. Cawley %B Active Learning and Experimental Design workshop In conjunction with AISTATS 2010 %C Proceedings of Machine Learning Research %D 2011 %E Isabelle Guyon %E Gavin Cawley %E Gideon Dror %E Vincent Lemaire %E Alexander Statnikov %F pmlr-v16-cawley11a %I PMLR %J Proceedings of Machine Learning Research %P 47--57 %U http://proceedings.mlr.press %V 16 %W PMLR %X In many potential applications of machine learning, unlabelled data are abundantly available at low cost, but there is a paucity of labelled data, and labeling unlabelled examples is expensive and/or time-consuming. This motivates the development of active learning methods, that seek to direct the collection of labelled examples such that the greatest performance gains can be achieved using the smallest quantity of labelled data. In this paper, we describe some simple pool-based active learning strategies, based on optimally regularised linear [kernel] ridge regression, providing a set of baseline submissions for the Active Learning Challenge. A simple random strategy, where unlabelled patterns are submitted to the oracle purely at random, is found to be surprisingly effective, being competitive with more complex approaches.
RIS
TY - CPAPER TI - Baseline Methods for Active Learning AU - Gavin C. Cawley BT - Active Learning and Experimental Design workshop In conjunction with AISTATS 2010 PY - 2011/04/21 DA - 2011/04/21 ED - Isabelle Guyon ED - Gavin Cawley ED - Gideon Dror ED - Vincent Lemaire ED - Alexander Statnikov ID - pmlr-v16-cawley11a PB - PMLR SP - 47 DP - PMLR EP - 57 L1 - http://proceedings.mlr.press/v16/cawley11a/cawley11a.pdf UR - http://proceedings.mlr.press/v16/cawley11a.html AB - In many potential applications of machine learning, unlabelled data are abundantly available at low cost, but there is a paucity of labelled data, and labeling unlabelled examples is expensive and/or time-consuming. This motivates the development of active learning methods, that seek to direct the collection of labelled examples such that the greatest performance gains can be achieved using the smallest quantity of labelled data. In this paper, we describe some simple pool-based active learning strategies, based on optimally regularised linear [kernel] ridge regression, providing a set of baseline submissions for the Active Learning Challenge. A simple random strategy, where unlabelled patterns are submitted to the oracle purely at random, is found to be surprisingly effective, being competitive with more complex approaches. ER -
APA
Cawley, G.C.. (2011). Baseline Methods for Active Learning. Active Learning and Experimental Design workshop In conjunction with AISTATS 2010, in PMLR 16:47-57

Related Material