Active Learning for Cost-Sensitive Classification

被引:0
|
作者
Krishnamurthy, Akshay [1 ]
Agarwal, Alekh [1 ]
Huang, Tzu-Kuo [2 ]
Daume, Hal, III [1 ]
Langford, John [1 ]
机构
[1] Microsoft Res, New York, NY 10011 USA
[2] Uber Adv Technol Ctr, Pittsburgh, PA 15201 USA
关键词
Active Learning; Cost-sensitive Learning; Structured Prediction; Statistical Learning Theory; Oracle-based Algorithms;
D O I
暂无
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
We design an active learning algorithm for cost-sensitive multiclass classification: problems where different errors have different costs. Our algorithm, COAL, makes predictions by regressing to each label's cost and predicting the smallest. On a new example, it uses a set of regressors that perform well on past data to estimate possible costs for each label. It queries only the labels that could be the best, ignoring the sure losers. We prove COAL can be efficiently implemented for any regression family that admits squared loss optimization; it also enjoys strong guarantees with respect to predictive performance and labeling effort. We empirically compare COAL to passive learning and several active learning baselines, showing significant improvements in labeling effort and test cost on real-world datasets.
引用
收藏
页数:50
相关论文
共 50 条
  • [1] Active Learning for Cost-Sensitive Classification
    Krishnamurthy, Akshay
    Agarwal, Alekh
    Huang, Tzu-Kuo
    Daume, Hal, III
    Langford, John
    [J]. INTERNATIONAL CONFERENCE ON MACHINE LEARNING, VOL 70, 2017, 70
  • [2] Active learning for cost-sensitive classification
    Krishnamurthy, Akshay
    Agarwal, Alekh
    Huang, Tzu-Kuo
    Daumé Iii, Hal
    Langford, John
    [J]. Journal of Machine Learning Research, 2019, 20
  • [3] Active Cost-Sensitive Learning
    Margineantu, Dragos D.
    [J]. 19TH INTERNATIONAL JOINT CONFERENCE ON ARTIFICIAL INTELLIGENCE (IJCAI-05), 2005, : 1622 - 1623
  • [4] Learning cost-sensitive active classifiers
    Greiner, R
    Grove, AJ
    Roth, D
    [J]. ARTIFICIAL INTELLIGENCE, 2002, 139 (02) : 137 - 174
  • [5] Active Learning for Cost-Sensitive Classification Using Logistic Regression Model
    Zhou, Siyuan
    Zhang, Ya
    [J]. PROCEEDINGS OF 2016 IEEE INTERNATIONAL CONFERENCE ON BIG DATA ANALYSIS (ICBDA), 2016, : 284 - 287
  • [6] Active Learning for Multiclass Cost-Sensitive Classification Using Probabilistic Models
    Chen, Po-Lung
    Lin, Hsuan-Tien
    [J]. 2013 CONFERENCE ON TECHNOLOGIES AND APPLICATIONS OF ARTIFICIAL INTELLIGENCE (TAAI), 2013, : 13 - 18
  • [7] Cost-Sensitive Active Visual Category Learning
    Vijayanarasimhan, Sudheendra
    Grauman, Kristen
    [J]. INTERNATIONAL JOURNAL OF COMPUTER VISION, 2011, 91 (01) : 24 - 44
  • [8] Cost-Sensitive Active Learning for Incomplete Data
    Wang, Min
    Yang, Chunyu
    Zhao, Fei
    Min, Fan
    Wang, Xizhao
    [J]. IEEE TRANSACTIONS ON SYSTEMS MAN CYBERNETICS-SYSTEMS, 2023, 53 (01): : 405 - 416
  • [9] Cost-Sensitive Active Visual Category Learning
    Sudheendra Vijayanarasimhan
    Kristen Grauman
    [J]. International Journal of Computer Vision, 2011, 91 : 24 - 44
  • [10] Cost-sensitive decision tree learning for forensic classification
    Davis, Jason V.
    Ha, Jungwoo
    Rossbach, Christopher J.
    Ramadan, Hany E.
    Witchel, Emmett
    [J]. MACHINE LEARNING: ECML 2006, PROCEEDINGS, 2006, 4212 : 622 - 629