Journal of Hebei University(Natural Science Edition) >

2017 , Vol. 37 >Issue 2: 216 - 224

DOI: https://doi.org/10.3969/j.issn.1000-1565.2017.02.017

Recent advances in active learning algorithms

  • YANG Wenzhu ,
  • TIAN Xiaoxiao ,
  • WANG Sile ,
  • ZHANG Xizhong
Expand
  • 1.School of Computer Science and Technology, Hebei University, Baoding 071002, China; 2.Institute of Information Technology, Baoding Education Examinations Authority, Baoding 071000, China

Received date: 2016-10-11

  Online published: 2017-03-25

Fold

Abstract

Active learning mainly aims at reducing the cost of manual annotation without decreasing the accuracy of the classifier.Active learning algorithm gets high quality training sample set by selecting the informative unlabeled samples which are labeled by domain experts later.The selected sample set is used to train the classifier.This improves the generalization ability of trained classifier while minimizes the cost of the labeling.Firstly,the recent advances in the three key steps in active learning algorithm was summarized,including:1)the method for constructing the initial training sample set and its improvement;2)the sample selection strategy and its improvement;3)the termination condition and its improvement.Then,the problems in active learning were analyzed and the corresponding countermeasures were presented.Finally,the future works in active learning were addressed.

Key words： active learning; initial training sample set; sample selection strategy; termination condition

Cite this article

YANG Wenzhu , TIAN Xiaoxiao , WANG Sile , ZHANG Xizhong . Recent advances in active learning algorithms[J]. Journal of Hebei University(Natural Science Edition), 2017 , 37(2) : 216 -224 . DOI: 10.3969/j.issn.1000-1565.2017.02.017

References

[1] 何清,李宁,罗文娟,等.大数据下的机器学习算法综述[J].模式识别与人工智能,2014,27(4):327-334. HE Q,LI N,LUO W J,et al.A survey of machine learning algorithms for Big Data[J].Pattern Recognition and Artificial Intellegence, 2014,27(4):327-334.
[2] KAREM F,DHIBI M,MARTIN A.Combination of supervised and unsupervised classification using the theory of belief functions[M].Belief Functions:Theory and Applications,Springer Berlin Heidelberg,2012:85-92.DOI:10.1007/978-3-642-29461-7_10.
[3] 翟俊海,张素芳,徐正夫,等.粗糙集与决策树比较研究[J].河北大学学报(自然科学版),2012,32(4):421-428. ZHAI J H,ZHANG S F,XU Z F,et al.Comparative study on rough sets and decision trees[J].Journal of Hebei University(Natural Science Edition),2012,32(4):421-428.
[4] FORESTIER G,WEMMERT C.Semi-supervised learning using multiple clusterings with limited labeled data[J].Information Sciences,2016,361:48-65.
[5] ZHU J,WANG H,TSOU B K,et al.Active learning with sampling by uncertainty and density for data annotations[J].IEEE Transactions on Audio Speech & Language Processing,2010,18(6):1323-1331.DOI:10.1109/TASL.2009.2033421
[6] 徐艳.基于主动学习的图像标注方法研究[D].辽宁:辽宁工业大学,2014. XU Y.Research on image annotation based active learning[D].Liaoning:Liaoning University of Technology,2014.
[7] 金良,曹永锋,苏彩霞,等.基于HS样例选择和BvSB反馈的多类图像分类[J].贵州师范大学学报(自然科学版),2014,32(04):56-61.DOI:10.16614/j.cnki.issn1004-5570.2014.04.013. JIN L,CAO Y F,SU C X,et al.Multi-class image classification based on HS sample selection and BvSB feedback[J].Guizhou Normal University(Natural Science Edition),2014,32(04):56-61.DOI:10.16614/j.cnki.issn1004-5570.2014.04.013.
[8] 赵秋焕.两种主动学习方法[D].保定:河北大学,2010. ZHAO Q H.Two kinds of active learning methods[D].Baoding:Hebei University,2010.
[9] 龙军,殷建平,祝恩,等.主动学习研究综述[J].计算机研究与发展,2008,45(s1):300-304. LONG J,YIN J P,ZHU E,et al.A survey of active learning[J].Journal of Computer Research and Development,2008,45(s1):300-304.
[10] CAMPBELLl C,CRISTIANINI N,SMOLA A.A query learning with large margin classifiers[Z].The 17th Int l Conf on Machine Learning,San Francisco:Morgan Kaufmann,2000.
[11] 韩冰,高新波,姬红兵.一种基于选择性集成SVM的新闻音频自动分类方法[J].模式识别与人工智能,2006,19(5):634-639. HAN B,GAO X B,JI H B.Automatic news audio classification method based on selective ensemble SVMS[J].PR&AI,2006,19(5):634-639.
[12] 刘康,钱旭,王自强.主动学习算法综述[J].计算机工程与应用,2012,48(34):1-4.DOI:10.3778/j.issn.1002-8331.1205-0149. LIU K,QIAN X,WANG Z Q.Overview of active learning algorithms[J].Computer Engineering and Applications,2012,48(34):1-4.DOI:10.3778/j.issn.1002-8331.1205-0149.
[13] ZHAO Y,XU C,CAO Y.Research on query-by-committee method of active learning and application[J].Computer Engineering,2006,4093(24):985-991.DOI:10.1007/11811305_107.
[14] 刘峰涛.基于样例池类标改变率的主动学习算法终止准则研究[D].保定:河北大学,2011. LIU F T.Research on change-rate-based stop criteria of active learning algorithms[D].Baoding:Hebei Vniversity,2011.
[15] ZHU J,WANG H,HOVY E,et al.Confidence-based stopping criteria for active learning for data annotation[J].Acm Transactions on Speech & Language Processing,2010,6(3):1-24.DOI:10.1145/1753783.1753784.
[16] WANG X Z,DONG L C,YAN J H.Maximum ambiguity-based sample selection in fuzzy decision tree induction[J].IEEE Transactions on Knowledge & Data Engineering,2012,24(8):1491-1505.DOI:10.1109/TKDE.2011.67.
[17] 白龙飞.基于支持向量机的主动学习方法研究[D].山西:山西大学,2012. BAI L F.Research on active learning approach based on support vector machines[D].Shanxi:University of Shanxi,2012.
[18] LIU J,YU H,YANG W,et al.Combining active learning and semi-supervised learning based on extreme learning machine for multi-class image classification[M] // Intelligence Science and Big DataEngineering,Image and Video Data Engineering.Springer International Publishing,2015.DOI:2015.10.1007/978-3-319-23989-7_18.
[19] JOSHI A J,PORIKLI F,PAPANIKOLOPOULOS N.Multi-class active learning for image classification[Z].Computer Society Conference on Computer Vision and Pattern Recognition,Miami,FL,2009.
[20] 王珍钰.基于不确定性的主动学习算法研究[D].保定:河北大学,2011. WANG Z Y.Study of active learning algorithms based on uncertainty[D].Baoding:Hebei University,2011.
[21] 陈荣,曹永锋,孙洪.基于主动学习和半监督学习的多类图像分类[J].自动化学报,2011,37(8):954-962.DOI:10.3724/SP.J.1004.2011.00954. CHEN R,CAO Y F,SUN H.Multi-class image classification with active learning and semi-supervised learning[J].Acta Automatica Sinica,2011,37(8):954-962.DOI:10.3724/SP.J.1004.2011.00954.
[22] HUANG S J,JIN R,ZHOU Z H.Active learning by querying infor-mative and representative examples[Z].The 24th Annual Conference on Neural Information Processing Systems,Va-ncouver,British Columbia,Canada:NIPS,2010.DOI:10.1109/TPAMI.2014.2307881.
[23] FRIEDMAN A,STEINBERG D,PIZARRO O,et al.Active learning using a Variational Dirichlet Process model for pre-clustering and classification of underwater stereo imagery[C] //Ieee/rsj International Conference on Intelligent Robots and Systems,Ieee/rsj International Conference on Intelligent Robots and Systems,2011:1533-1539.DOI:10.1109/IROS.2011.6095178.
[24] DONMWZ P,CARBONELL J G,BENNETT P N.Dual strategy active learning[C] //Proceedings of the 18th European Conference on Machine Learning.Springer-Verlag:Springer,2007:116-127.DOI:10.1007/978-3-540-74958-5_14.
[25] 胡正平,高文涛, 万春艳.基于样本不确定性和代表性相结合的可控主动学习算法研究[J].燕山大学学报,2009,33(4):341-346. HU Z P,GAO W T,WAN C Y.Research on controlled active learning algorithm based on the combination of sample uncertainty and representation[J].Journal of Yanshan University,2009,33(4):341-346.
[26] 曹永锋,陈荣,孙洪.基于BvSBHC的主动学习多类分类算法[J].计算机科学,2013,40(8):309-312. CAO Y F,CHEN R,SUN H.Multi-class image classification with best vs.second-best active learning and hierarchical clustering[J].Computer Science,2013,40(8):309-312.
[27] 吴伟宁,刘扬,郭茂祖,等.基于采样策略的主动学习算法研究进展[J].计算机研究与发展,2012,49(6):1162-1173. WU W N,LIU Y,GUO M Z,et al.Advances in active learning algorithms based on sampling strategy[J].Journal of Computer Research and Development,2012,49(6):1162-1173.
[28] 陈昀,毕海岩.基于多特征融合的中文评论情感分类算法[J].河北大学学报(自然科学版),2015,35(6):651-656.DOI:10.3969/j.issn1000-1565.2015.06.016. CHEN Y,BI H Y.A sentiment classification algorithm of Chinese comments based on multi features fusion[J].Journal of Hebei University(Natural Science Edition),2015,35(6):651-656.DOI:10.3969/j.issn1000-1565.2015.06.016.
[29] 韩光,赵春霞,胡雪蕾.一种新的SVM主动学习算法及其在障碍物检测中的应用[J].计算机研究与发展,2009,46(11):1934-1941. HAN G,ZHAN C X,HU X L.An SVM active learning algorithm and its application in obstacle detection[J].Journal of Computer Research and Development,2009,46(11):1934-1941.
[30] 鲍翠梅.基于主动学习的加权支持向量机的分类[J].计算机工程与设计,2009,30(4):966-970.DOI:10.16208/j.issn1000-7024.2009.04.071. BAO C M.Classification of weighted support vector machine based on active learning[J].Computer Engineering and Design,2009,30(4):966-970.DOI:10.16208/j.issn1000-7024.2009.04.071.
[31] 梁延峰.基于专家委员会的主动学习算法研究[D].青岛:中国海洋大学,2010. LIANG Y F.Research of query-by-committee method of active learning[D].Qingdao:Chinese Marine University,2010.
[32] 高成,陈秀新,于重重,等.基于主动学习的图半监督分类算法[J].计算机工程与设计,2015(7):1871-1875.DOI:10.16208/j.issn1000-7024.2015.07.037. GAO C,CHEN X X,YU C C,et al.Graph-based semi-supervised classification algorithm based on active learning[J].Computer engineering and design,2015(7):1871-1875 DOI:10.16208/j.issn1000-7024.2015.07.037.
[33] 白龙飞,王文剑,郭虎升.一种新的支持向量机主动学习策略[J].南京大学学报(自然科学),2012,48(2):182-189.DOI:10.13232/j.cnki.jnju.2012.02.008. BAI L F,WANG W J,GUO H S.A novel support vector machine active learning strategy[J].Journal of Nanjingl University(Natural Sciences),2012,48(2):182-189.DOI:10.13232/j.cnki.jnju.2012.02.008.
[34] 徐美香,孙福明,李豪杰.主动学习的多标签图像在线分类[J].中国图象图形学报,2015,20(2):237-244.DOI:10.11834 /jig.20150210. XU M X,SUN F M,LI H J.Online multi-label image classification with active learning[J].Journal of image and Graphics,2015,20(2):237-244.DOI:10.11834 /jig.20150210.
[35] 李海峰,李纯果.深度学习结构和算法比较分析[J].河北大学学报(自然科学版),2012,32(5):538-544. LI H F,LI C G.Note on deep architecture and deep learning algorithms[J].Journal of Hebei University(Natural Science Edition),2012,32(5):538-544.
Options
Outlines

/

〈 〉