Student or educator? Sign up with your school email and get 6 months of Pro free.

The DOI matches its Crossref record; 63 of the 63 references it cites were found in a registry or an index.

A brief introduction to weakly supervised learning

Open access
Zhou, Zhi-Hua·2018·National Science Review

Abstract

Supervised learning techniques construct predictive models by learning from a large number of training examples, where each training example has a label indicating its ground-truth output. Though current techniques have achieved great success, it is noteworthy that in many tasks it is difficult to get strong supervision information like fully ground-truth labels due to the high cost of the data-labeling process. Thus, it is desirable for machine-learning techniques to work with weak supervision. This article reviews some research progress of weakly supervised learning, focusing on three typical types of weak supervision: incomplete supervision, where only a subset of training data is given with labels; inexact supervision, where the training data are given with only coarse-grained labels; and inaccurate supervision, where the given labels are not always ground-truth.

Overview

Integrity signals

Registry

Crossref · 10.1093/nsr/nwx106

Cites

63 of 63 references verified0 citation errors · 0 not checkable

Statistics

Recomputing p-values needs the paper’s full text.

Verification

Claims the paper makes

We couldn’t retrieve this paper’s full text, so its claims and self-check weren’t run. The registry and citation checks still apply.

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Citations

Cites

Show all 63 references

Research works printed but not linked: 48

  1. Settles, 2010, Active learning literature survey
  2. Lewis, 1994, A sequential algorithm for training text classifiers, 3
  3. Seung, 1992, Query by committee, 287
  4. Abe, 1998, Query learning strategies using boosting and bagging, 1
  5. Nguyen, 2004, 21st International Conference on Machine Learning, Banff, Canada, 623
  6. Dasgupta, 2005, Analysis of perceptron-based active learning, 249
  7. Dasgupta, 2005, Analysis of a greedy active learning strategy, 337
  8. Balcan, 2007, Margin based active learning, 35
  9. Hanneke, 2009, Adaptive rates of convergence in active learning
  10. Wang, 2010, Multi-view active learning in the non-realizable case, 2388
  11. Miller, 1997, A mixture of experts classifier with learning based on both labelled and unlabelled data, 571
  12. Fujino, 2005, A hybrid generative/discriminative approach to semi-supervised classifier design, 764
  13. Zhu, 2003, Semi-supervised learning using Gaussian fields and harmonic functions, 912
  14. Zhou, 2004, Learning with local and global consistency, 321
  15. Hein, 2007, Manifold denoising, 561
  16. Joachims, 1999, Transductive inference for text classification using support vector machines, 200
  17. Chapelle, 2005, Semi-supervised learning by low density separation, 57
  18. Blum, 1998, Combining labeled and unlabeled data with co-training, 92
  19. Cozman, 2002, Unlabeled data can degrade classification performance of generative classifiers, 327
  20. Zhou, 2007, On the relation between multi-instance learning and semi-supervised learning, 1167
  21. Zhou, 2009, Multi-instance learning by treating instances as non-i.i.d. samples, 1249
  22. Zhang, 2002, Content-based image retrieval using multiple-instance learning, 682
  23. Andrews, 2003, Support vector machines for multiple-instance learning, 561
  24. Settles, 2008, Multiple-instance active learning, 1289
  25. Fung, 2007, Multiple instance learning for computer aided diagnosis, 425
  26. Liu, 2012, Key instance detection in multi-instance learning, 253
  27. Xu, 2004, Logistic regression and boosting for labeled bags of instances, 272
  28. Weidmann, 2003, A two-level learning method for generalized multi-instance problem, 468
  29. Sabato, 2009, Homogenous multi-instance learning with arbitrary dependence
  30. Gao, 2016, Risk minimization in the presence of label noise, 1575
  31. Snow, 2008, Cheap and fast - but is it good? Evaluating non-expert annotations for natural language tasks, 254
  32. Whitehill, 2009, Whose vote should count more: optimal integration of labels from labelers of unknown expertise, 2035
  33. Dekel, 2009, Good learners for evil teachers, 233
  34. Urner, 2012, Learning from weak teachers, 1252
  35. Wang, 2016, Cost-saving effect of crowdsourcing learning, 2111
  36. Karger, 2011, Iterative learning for reliable crowdsourcing systems, 1953
  37. Tran-Thanh, 2013, Efficient budget allocation with accuracy guarantees for crowdsourcing classification tasks, 901
  38. Ho, 2013, Adaptive task assignment for crowdsourced classification, 534
  39. Chen, 2013, Optimistic knowledge gradient policy for optimal budget allocation in crowdsourcing, 64
  40. Zhong, 2015, Active learning from crowds with unsure option, 1061
  41. Ding, 2016, Crowdsourcing with unsure opinion
  42. Shah, 2015, Double or nothing: multiplicative incentive mechanisms for crowdsourcing, 1
  43. Yan, 2011, Active learning from crowds, 1161
  44. Garcia-Garcia, 2011, Degrees of supervision
  45. Sun, 2010, Multi-label learning with weak label, 593
  46. Li, 2013, Active learning with multi-label SVM classification, 1479
  47. Qi, 2008, Two-dimensional active learning for image classification
  48. Huang, Multi-label active learning: query type matters, 946

Cited by

See all 1,899

Metadata

Authors
Zhou, Z.-H.
Affiliations
Nanjing University (CN)
Corresponding author
Zhou, Zhi-Hua
Venue
National Science Review
Published
2017-08-25
Volume
5
Issue
1
Pages
44-53
ISSN
  • 2053-714X (ISSN-L)
  • 2095-5138
OpenAlex
W2746791238
Cited by
1,899
References
63
Citations by year
2026: 173 · 2025: 228 · 2024: 268 · 2023: 319 · 2022: 264 · 2021: 283 · 2020: 193 · 2019: 123 · 2018: 45 · 2017: 3
FWCI
58.56
Type
journal article
Language
en
Open access
Licence
cc-by
Locations
National Science Review (publishedVersion, cc-by)
Funders
National Natural Science Foundation of China (61333014)
Topics
  • Physical Sciences › Computer Science › Artificial Intelligence › Machine Learning and Algorithms
  • Physical Sciences › Computer Science › Artificial Intelligence › Machine Learning and Data Classification
  • Physical Sciences › Computer Science › Artificial Intelligence › Domain Adaptation and Few-Shot Learning
Keywords
weakly supervised learning, label noise, data labeling cost, training data labeling

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: A brief introduction to weakly supervised learning. https://app.cochat.ai/papers/10.1093/nsr/nwx106
The paper
Zhou, Z.-H. (2018). A brief introduction to weakly supervised learning. National Science Review. https://doi.org/10.1093/nsr/nwx106

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.