Learning from Multiple Noisy Partial Labelers

被引:0
|
作者
Yu, Peilin [1 ]
Ding, Tiffany [2 ]
Bach, Stephen H. [1 ]
机构
[1] Brown Univ, Providence, RI 02912 USA
[2] Univ Calif Berkeley, Berkeley, CA USA
关键词
D O I
暂无
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Programmatic weak supervision creates models without hand-labeled training data by combining the outputs of heuristic labelers. Existing frameworks make the restrictive assumption that labelers output a single class label. Enabling users to create partial labelers that output subsets of possible class labels would greatly expand the expressivity of programmatic weak supervision. We introduce this capability by defining a probabilistic generative model that can estimate the underlying accuracies of multiple noisy partial labelers without ground truth labels. We show how to scale up learning, for example learning on 100k examples in one minute, a 300x speed up compared to a naive implementation. We also prove that this class of models is generically identifiable up to label swapping under mild conditions. We evaluate our framework on three text classification and six object classification tasks. On text tasks, adding partial labels increases average accuracy by 8.6 percentage points. On image tasks, we show that partial labels allow us to approach some zero-shot object classification problems with programmatic weak supervision by using class attributes as partial labelers. On these tasks, our framework has accuracy comparable to recent embedding-based zero-shot learning methods, while using only pre-trained attribute detectors.
引用
收藏
页数:24
相关论文
共 50 条
  • [31] The speed of learning in noisy games: Partial reinforcement and the sustainability of cooperation
    Bereby-Meyer, Yoella
    Roth, Alvin E.
    AMERICAN ECONOMIC REVIEW, 2006, 96 (04): : 1029 - 1042
  • [32] A Learning Framework for Cognitive Interference Networks with Partial and Noisy Observations
    Levorato, Marco
    Firouzabadi, Sina
    Goldsmith, Andrea
    IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS, 2012, 11 (09) : 3101 - 3111
  • [33] Partial multi-label learning with noisy side information
    Sun, Lijuan
    Feng, Songhe
    Lyu, Gengyu
    Zhang, Hua
    Dai, Guojun
    KNOWLEDGE AND INFORMATION SYSTEMS, 2021, 63 (02) : 541 - 564
  • [34] Cognitive Interference Networks with Partial and Noisy Observations: a Learning Framework
    Levorato, Marco
    Firouzabadi, Sina
    Goldsmith, Andrea
    2011 IEEE GLOBAL TELECOMMUNICATIONS CONFERENCE (GLOBECOM 2011), 2011,
  • [35] Partial Multi-Label Learning with Noisy Label Identification
    Xie, Ming-Kun
    Huang, Sheng-Jun
    THIRTY-FOURTH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE, THE THIRTY-SECOND INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE AND THE TENTH AAAI SYMPOSIUM ON EDUCATIONAL ADVANCES IN ARTIFICIAL INTELLIGENCE, 2020, 34 : 6454 - 6461
  • [36] Combining machine learning and data assimilation to forecast dynamical systems from noisy partial observations
    Gottwald, Georg A.
    Reich, Sebastian
    CHAOS, 2021, 31 (10)
  • [37] Robust Equivariant Imaging: a fully unsupervised framework for learning to image from noisy and partial measurements
    Chen, Dongdong
    Tachella, Julian
    Davies, Mike E.
    2022 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR 2022), 2022, : 5637 - 5646
  • [38] A probabilistic model of active learning with multiple noisy oracles
    Wu, Weining
    Liu, Yang
    Guo, Maozu
    Wang, Chunyu
    Liu, Xiaoyan
    NEUROCOMPUTING, 2013, 118 : 253 - 262
  • [39] DSAL: Deeply Supervised Active Learning From Strong and Weak Labelers for Biomedical Image Segmentation
    Zhao, Ziyuan
    Zeng, Zeng
    Xu, Kaixin
    Chen, Cen
    Guan, Cuntai
    IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS, 2021, 25 (10) : 3744 - 3751
  • [40] Learning to Rank From a Noisy Crowd
    Kumar, Abhimanu
    Lease, Matthew
    PROCEEDINGS OF THE 34TH INTERNATIONAL ACM SIGIR CONFERENCE ON RESEARCH AND DEVELOPMENT IN INFORMATION RETRIEVAL (SIGIR'11), 2011, : 1221 - 1222