Cost-sensitive active learning through farthest distance sum sampling

doi:10.11772/j.issn.1001-9081.2019020763

Journal of Computer Applications ›› 2019, Vol. 39 ›› Issue (9): 2499-2504.DOI: 10.11772/j.issn.1001-9081.2019020763

• Artificial intelligence • Previous Articles Next Articles

Cost-sensitive active learning through farthest distance sum sampling

REN Jie1, MIN Fan1, WANG Min2

1. School of Computer Science, Southwest Petroleum University, Chengdu Sichuan 610500, China;
2. School of Electrical Engineering and Information, Southwest Petroleum University, Chengdu Sichuan 610500, China

Received:2019-03-22 Revised:2019-05-06 Online:2019-09-10 Published:2019-06-03
Supported by:
This work is partially supported by the Scientific Innovation Group for Youths of Sichuan Province (2019JDTD0017), the Applied Basic Research Project of Sichuan Province (2017JY0190).

基于最远总距离采样的代价敏感主动学习

任杰¹, 闵帆¹, 汪敏²

1. 西南石油大学计算机科学学院, 成都 610500;
2. 西南石油大学电气信息学院, 成都 610500

通讯作者: 闵帆
作者简介:任杰(1996-),男,山西忻州人,硕士研究生,主要研究方向:主动学习;闵帆(1973-),男,重庆人,教授,博士,CCF会员,主要研究方向:粒计算、推荐系统、主动学习;汪敏(1980-),女,湖南邵阳人,副教授,硕士,CCF会员,主要研究方向:数据挖掘、主动学习。
基金资助:
四川省青年科技创新团队专项（2019JDTD0017）；四川省应用基础研究项目（2019JDTD0017）。

Abstract

Abstract:

Active learning aims to reduce expert labeling through man-machine interaction, while cost-sensitive active learning focuses on balancing labeling and misclassification costs. Based on Three-Way Decision (3WD) methodology and Label Uniform Distribution (LUD) model, a Cost-sensitive Active learning through the Farthest distance sum Sampling (CAFS) algorithm was proposed. Firstly, the farthest total distance sampling strategy was designed to query the labels of representative samples. Secondly, LUD model and cost function were used to calculate the expected sampling number. Finally, k-Means algorithm was employed to split blocks obtained different labels. In CAFS, 3WD methodology was adopted in the iterative process of label query, instance prediction, and block splitting, until all instances were processed. The learning process was controlled by the cost minimization objective. Results on 9 public datasets show that CAFS has lower average cost compared with 11 mainstream algorithms.

Key words: active learning, k-Means clustering, label uniform distribution, Three-Way Decision (3WD)

摘要：

主动学习旨在通过人机交互减少专家标注，代价敏感主动学习则致力于平衡标注与误分类代价。基于三支决策（3WD）和标签均匀分布（LUD）模型，提出一种基于最远总距离采样的代价敏感主动学习算法（CAFS）。首先，设计了最远总距离采样策略，以查询代表性样本的标签；其次，利用了LUD模型和代价函数，计算期望采样数目；最后，使用了k-Means聚类技术分裂已获得不同标签的块。CAFS算法利用三支决策思想迭代地进行标签查询、实例预测和块分裂，直至处理完所有实例。学习过程在代价最小化目标的控制下进行。在9个公开数据上比较，CAFS比11个主流的算法具有更低的平均代价。

关键词: 主动学习, k-Means聚类, 标签均匀分布, 三支决策

CLC Number:

TP181

REN Jie, MIN Fan, WANG Min. Cost-sensitive active learning through farthest distance sum sampling[J]. Journal of Computer Applications, 2019, 39(9): 2499-2504.

任杰, 闵帆, 汪敏. 基于最远总距离采样的代价敏感主动学习[J]. 计算机应用, 2019, 39(9): 2499-2504.

References

[1] SETTLES B. Active Learning[M]. San Rafael, CA:Morgan and Claypool Publishers, 2012:1-114.
[2] ZHU X, GOLDBERG A B. Introduction to Semi-Supervised Learning[M]. San Rafael, CA:Morgan and Claypool Publishers, 2009:130.
[3] SEUNG H S, OPPER M, SOMPOLINSKY H. Query by committee[C]//COLT 1992:Proceedings of the 5th Annual ACM Conference on Computational Learning Theory. New York:ACM, 1992:287-294.
[4] COHN D A, GHAHRAMANI Z, JORDAN M I, et al. Active learning with statistical models[J]. Journal of Artificial Intelligence Research, 1996, 4(1):129-145.
[5] WANG M, MIN F, ZHANG Z H, et al. Active learning through density clustering[J]. Expert Systems with Applications, 2017, 85:305-317.
[6] TONG S, KOLLER D. Support vector machine active learning with applications to text classification[J]. Journal of Machine Learning Research, 2001, 2(1):45-66.
[7] THOMPSON C A. Active learning for natural language parsing and information extraction[C]//ICML 1999:Proceeding of the 16th International Conference on Machine Learning. San Francisco, CA:Morgan Kaufmann Publishers, 1999:406-414.
[8] ZHANG C, CHEN T. An active learning framework for content-based information retrieval[J]. IEEE Transactions on Multimedia, 2002, 4(2):260-268.
[9] YU D, VARADARAJAN B, DENG L, et al. Active learning and semi-supervised learning for speech recognition:a unified framework using the global entropy reduction maximization criterion[J]. Computer Speech and Language, 2010, 24(3):433-444.
[10] MARGINEANTU D D. Active cost-sensitive learning[C]//IJCAI 2005:Proceedings of the 19th International Joint Conference on Artificial Intelligence. San Francisco, CA:Morgan Kaufmann Publishers, 2005:1622-1623.
[11] MIN F, LIU F L, WEN L Y, et al. Tri-partition cost-sensitive active learning through kNN[J]. Soft Computing, 2017, 23(5):1557-1572.
[12] WU Y X, MIN X Y, MIN F, et al. Cost-sensitive active learning with a label uniform distribution model[J]. International Journal of Approximate Reasoning, 2019, 105:49-65.
[13] YAO Y. Three-way decision:an interpretation of rules in rough set theory[C]//Proceedings of the 2009 International Conference on Rough Sets and Knowledge Technology, LNCS 5589. Berlin:Springer, 2009:642-649.
[14] 李华雄,周献中,黄兵,等.决策粗糙集与代价敏感分类[J].计算机科学与探索,2013,7(2):126-135.(LI H X, ZHOU X Z, HUANG B, et al. Decision-theoretic rough set and cost-sensitive classification[J]. Journal of Frontiers of Computer Science and Technology, 2013, 7(2):126-135.)
[15] 刘盾,李天瑞,李华雄.粗糙集理论:基于三支决策视角[J].南京大学学报(自然科学版),2013,49(5):574-581. (LIU D, LI T R, LI H X. Rough set theory:a three-way decisions perspective[J]. Journal of Nanjing University (Natural Science), 2013, 49(5):574-581)
[16] 杨习贝,杨静宇.邻域系统粗糙集模型[J].南京理工大学报,2012,36(2):291-295.(YANG X B, YANG J Y. Rough set model based on neighborhood system[J]. Journal of Nanjing University of Science and Technology, 2012, 36(2):291-295.)
[17] SETTLES B, CRAVEN M, Friedland L. Active learning with real annotation costs[EB/OL].[2018-12-13]. https://www.researchgate.net/publication/228770726_Active_learning_with_real_annotation_costs.
[18] LIU A, JUN G, GHOSH J. Spatially cost-sensitive active learning[C]//SDM 2009:Proceedings of the 2009 SIAM International Conference on Data Mining. Philadelphia, PA:SIAM, 2009:814-825.
[19] ZHAO P L, HOI S C H. Cost-sensitive online active learning with application to malicious URL detection[C]//KDD 2013:Proceedings of the 19th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. New York:ACM, 2013:919-927.
[20] CHEN P-L, LIN H-T. Active learning for multiclass cost-sensitive classification using probabilistic models[C]//TAAI 2013:Proceedings of the 2013 Conference on Technologies and Applications of Artificial Intelligence. Washington, DC:IEEE Computer Society, 2013:13-18.
[21] DEMIR B, MINELLO L, BRUZZONE L. Definition of effective training sets for supervised classification of remote sensing images by a novel cost-sensitive active learning method[J]. IEEE Transactions on Geoscience and Remote Sensing, 2014, 52(2):1272-1284.
[22] HUANG K-H, LIN H-T. A novel uncertainty sampling algorithm for cost-sensitive multiclass active learning[C]//ICDM 2016:Proceedings of the 2016 IEEE 16th International Conference on Data Ming. Piscataway, NJ:IEEE, 2016:925-930.
[23] BAHNSEN A C, AOUADA D, OTTERSTEN B. Example-dependent cost-sensitive logistic regression for credit scoring[C]//ICMLA 2014:Proceedings of the 2014 13th International Conference on Machine Learning and Application. Washington, DC:IEEE Computer Society, 2014:263-269.
[24] BAHNSEN A C, AOUADA D, OTTERSTEN B. Example-dependent cost-sensitive decision trees[J]. Expert Systems with Applications, 2015, 42(19):6609-6619.
[25] BAHNSEN A C, AOUADA D, OTTERSTEN B. Ensemble of example-dependent cost-sensitive decision trees[EB/OL].[2018-12-13]. https://arxiv.org/pdf/1505.04637v1.pdf.
[26] QUINLAN J R. Induction of decision trees[J]. Machine Learning, 1986, 1(1):81-106.
[27] LIAW A, WIENER M. Classification and regression by random forest[J]. R News, 2002, 2/3:18-22.
[28] CRISTIANINI N, SHAWE T J. An Introduction to Support Vector Machines and Other Kernel-Based Learning Methods[M]. Cambridge, Eng.:Cambridge University Press, 2000:46-71.

Cost-sensitive active learning through farthest distance sum sampling

基于最远总距离采样的代价敏感主动学习

PDF

Knowledge

Abstract

Cite this article

share this article

References

Related Articles 15

Recommended Articles

Metrics

[1]	ZHANG Shipeng, LI Yongzhong, DU Xiangtong. Intrusion detection model based on semi-supervised learning and three-way decision [J]. Journal of Computer Applications, 2021, 41(9): 2602-2608.
[2]	YI Dongyi, DENG Genqiang, DONG Chaoxiong, ZHU Miaomiao, LYU Zhouping, ZHU Suisong. Medical insurance fraud detection algorithm based on graph convolutional neural network [J]. Journal of Computer Applications, 2020, 40(5): 1272-1277.
[3]	WANG Min, WU Yubo, MIN Fan. Multi-category active learning algorithm based on multiple clustering algorithms and multivariate linear regression [J]. Journal of Computer Applications, 2020, 40(12): 3437-3444.
[4]	GONG Yanlu, LYU Jia. Co-training algorithm with combination of active learning and density peak clustering [J]. Journal of Computer Applications, 2019, 39(8): 2297-2301.
[5]	QIU Ningjia, CONG Lin, ZHOU Sicheng, WANG Peng, LI Yanfang. SVD-CNN barrage text classification algorithm combined with improved active learning [J]. Journal of Computer Applications, 2019, 39(3): 644-650.
[6]	CHAI Bianfang, LYU Feng, LI Wenbin, WANG Yao. Semi-supervised K-means clustering algorithm based on active learning priors [J]. Journal of Computer Applications, 2018, 38(11): 3139-3143.
[7]	ZHAI Junhai, ZHANG Sufang, WANG Cong, SHEN Chu, LIU Xiaomeng. Big data active learning based on MapReduce [J]. Journal of Computer Applications, 2018, 38(10): 2759-2763.
[8]	WANG Lili, FU Zhongliang, TAO Pan, ZHU Kai. TEE standard plane classification based on improved multi-class AdaBoost algorithm [J]. Journal of Computer Applications, 2017, 37(8): 2253-2257.
[9]	WANG Lili, FU Zhongliang, TAO Pan, HU Xin. Heart disease classification based on active imbalance multi-class AdaBoost algorithm [J]. Journal of Computer Applications, 2017, 37(7): 1994-1998.
[10]	ZHANG Liang, LUO Yimin, MA Hongchao, ZHANG Fan, HU Chuan. Hyperspectral remote sensing image classification based on active learning algorithm with unlabeled information [J]. Journal of Computer Applications, 2017, 37(6): 1768-1771.
[11]	ZHU Jiang, MING Yue, WANG Sen. Mechanism of security situation element acquisition based on deep auto-encoder network [J]. Journal of Computer Applications, 2017, 37(3): 771-776.
[12]	CHAI Bianfang, WANG Jianling, XU Jiwei, LI Wenbin. Active semi-supervised community detection method based on link model [J]. Journal of Computer Applications, 2017, 37(11): 3090-3094.
[13]	LI Yanling, YAN Yonghong. Weakly-supervised training method about Chinese spoken language understanding [J]. Journal of Computer Applications, 2015, 35(7): 1965-1968.
[14]	YANG Ju, LI Qingwen, YU Hualong. Stopping criterion of active learning for scenario of single-labeling mode [J]. Journal of Computer Applications, 2015, 35(12): 3472-3476.
[15]	HAO Zedong YU Songsong GUAN Jihong. Hyperspectral image classification based on active learning [J]. Journal of Computer Applications, 2013, 33(12): 3441-3443.