mlsanityData to 7 Oct 2026

Identifiability of Label Noise Transition Matrix

arXiv 2202.02016 ↗First posted 4 Feb 2022cs.LG stat.MLcited by 34

Not on the map. Why?

Abstract (arXiv)

The noise transition matrix plays a central role in the problem of learning with noisy labels. Among many other reasons, a large number of existing solutions rely on access to it. Identifying and estimating the transition matrix without ground truth labels is a critical and challenging task. When label noise transition depends on each instance, the problem of identifying the instance-dependent noise transition matrix becomes substantially more challenging. Despite recent works proposing solutions for learning from instance-dependent noisy labels, the field lacks a unified understanding of when such a problem remains identifiable. The goal of this paper is to characterize the identifiability of the label noise transition matrix. Building on Kruskal's identifiability results, we are able to show the necessity of multiple noisy labels in identifying the noise transition matrix for the generic case at the instance level. We further instantiate the results to explain the successes of the state-of-the-art solutions and how additional assumptions alleviated the requirement of multiple noisy labels. Our result also reveals that disentangled features are helpful in the above identification task and we provide empirical evidence.

Key facts. Cited by 34 papers in the map (data through 7 Oct 2026). Builds on 40 of its 70 references that are in the map. First posted on arXiv 4 Feb 2022.

Builds on

40 of 70 references.

References · 40

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2020A Simple Framework for Contrastive Learning of Visual Representationscited by 11,979
  3. 2018Representation Learning with Contrastive Predictive Codingcited by 7,662
  4. 2019Invariant Risk Minimizationcited by 1,832
  5. 2018Co-teaching: Robust Training of Deep Neural Networks with Extremely Noisy Labelscited by 1,053
  6. 2020DivideMix: Learning with Noisy Labels as Semi-supervised Learningcited by 691
  7. 2016Making Deep Neural Networks Robust to Label Noise: a Loss Correction Approachcited by 663
  8. 2017MentorNet: Learning Data-Driven Curriculum for Very Deep Neural Networks on Corrupted Labelscited by 652
  9. 2019Does Learning Require Memorization? A Short Tale about a Long Tailcited by 452
  10. 2019Confident Learning: Estimating Uncertainty in Dataset Labelscited by 411
  11. 2018Towards a Definition of Disentangled Representationscited by 294
  12. 2020Combating noisy labels by agreement: A joint training method with co-regularizationcited by 289
28 more
  1. 2021Learning with Noisy Labels Revisited: A Study Using Real-World Human Annotationscited by 253
  2. 2016Tensor Decomposition for Signal Processing and Machine Learningcited by 223
  3. 2019Are Anchor Points Really Indispensable in Label-Noise Learning?cited by 208
  4. 2019SELF: Learning to Filter Noisy Labels with Self-Ensemblingcited by 164
  5. 2020Learning with Instance-Dependent Label Noise: A Sample Sieve Approachcited by 158
  6. 2020Dual T: Reducing Estimation Error for Transition Matrix in Label-noise Learningcited by 142
  7. 2021Provably End-to-end Label-Noise Learning without Anchor Pointscited by 97
  8. 2020A Second-Order Approach to Learning with Instance-Dependent Label Noisecited by 90
  9. 2021Contrast to Divide: Self-Supervised Pre-Training for Learning with Noisy Labelscited by 69
  10. 2020Fair Classification with Group-Dependent Label Noisecited by 67
  11. 2021Clusterability as an Alternative to Anchor Points When Learning with Noisy Labelscited by 65
  12. 2014Spectral Methods meet EM: A Provably Optimal Algorithm for Crowdsourcingcited by 63
  13. 2020Confidence Scores Make Instance-dependent Label-noise Learning Possiblecited by 63
  14. 2021Learning Noise Transition Matrix from Only Noisy Labels via Total Variation Regularizationcited by 62
  15. 2017Learning with Bounded Instance- and Label-dependent Label Noisecited by 57
  16. 2020Deep k-NN for Noisy Labelscited by 55
  17. 2017Learning with Confident Examples: Rank Pruning for Robust Classification with Noisy Labelscited by 52
  18. 2021Instance-dependent Label-noise Learning under a Structural Causal Modelcited by 50
  19. 2021Self-Supervised Learning Disentangled Group Representation as Featurecited by 47
  20. 2020Part-dependent Label Noise: Towards Instance-dependent Label Noisecited by 40
  21. 2021Contrastive Learning Improves Model Robustness Under Label Noisecited by 36
  22. 2018Improving Generalization for Abstract Reasoning Tasks Using Disentangled Feature Representationscited by 28
  23. 2022Beyond Images: Label Noise Transition Matrix Estimation for Tasks with Lower-Quality Featurescited by 27
  24. 2021Understanding Instance-Level Label Noise: Disparate Impacts and Treatmentscited by 22
  25. 2021Can Less be More? When Increasing-to-Balancing Label Noise Rates Considered Beneficialcited by 13
  26. 2017Blind Multiclass Ensemble Classificationcited by 8
  27. 2021Learning with Group Noisecited by 5
  28. 2018Surrogate Scoring Rulescited by 4

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2022: 1; 2023: 14; 2024: 8; 2025: 7; 2026: 4.

34 papers.

Sort:

By area:

  1. 2023Unmasking and Improving Data Credibility: A Study with Datasets for Training Harmless Language Modelscited by 26
  2. 2024Improving Data Efficiency via Curating LLM-Driven Rating Systemscited by 26
  3. 2022Weak Proxies are Sufficient and Preferable for Fairness with Missing Sensitive Attributescited by 17
  4. 2023InstanT: Semi-supervised Learning with Instance-dependent Thresholdscited by 13
  5. 2023Transferring Annotator- and Instance-dependent Transition Matrix for Learning from Crowdscited by 11
  6. 2023Regularly Truncated M-estimators for Learning with Noisy Labelscited by 8
  7. 2024Tackling Noisy Labels with Network Parameter Additive Decompositioncited by 8
  8. 2024Learning to Complement and to Defer to Multiple Userscited by 7
  9. 2023Learning to Complement with Multiple Humanscited by 5
  10. 2023Instance-dependent Noisy-label Learning with Graphical Model Based Noise-rate Estimationcited by 4
  11. 2024Dirichlet-based Per-Sample Weighting by Transition Matrix for Noisy Label Learningcited by 4
  12. 2023Improve Noise Tolerance of Robust Loss via Noise-Awarenesscited by 3
22 more
  1. 2024Understanding and Mitigating Human-Labelling Errors in Supervised Contrastive Learningcited by 3
  2. 2025When Human Preferences Flip: An Instance-Dependent Robust Loss for RLHFcited by 3
  3. 2023LNL+K: Enhancing Learning with Noisy Labels Through Noise Source Knowledge Integrationcited by 2
  4. 2023Elucidating and Overcoming the Challenges of Label Noise in Supervised Contrastive Learningcited by 2
  5. 2025Noise-Aware Generalization: Robustness to In-Domain Noise and Out-of-Domain Generalizationcited by 2
  6. 2025Foundation Model Priors Enhance Object Focus in Feature Space for Source-Free Object Detectioncited by 2
  7. 2026L2CU: Learning to Complement Unseen Userscited by 2
  8. 2023Towards the Identifiability in Noisy Label Learning: A Multinomial Mixture Modelling Approachcited by 1
  9. 2023Human-annotated label noise and their impact on ConvNets for remote sensing image scene classificationcited by 1
  10. 2023Bridging Generative and Discriminative Noisy-Label Learning via Direction-Agnostic EM Formulationcited by 1
  11. 2024NoisyAG-News: A Benchmark for Addressing Instance-Dependent Noise in Text Classificationcited by 1
  12. 2024Mislabeled examples detection viewed as probing machine learning models: concepts, survey and extensive benchmarkcited by 1
  13. 2026Deconstructing the Failure of Ideal Noise Correction: A Three-Pillar Diagnosiscited by 1
  14. 2026MOLAR: Learning Multimodal Molecular Representations from Noisy Labelscited by 1
  15. 2023Multi-Label Noise Transition Matrix Estimation with Label Correlations: Theory and Algorithmcited by 0
  16. 2023Hypothesis Testing for Class-Conditional Noise Using Local Maximum Likelihoodcited by 0
  17. 2024EchoAlign: Bridging Generative and Discriminative Learning under Noisy Labelscited by 0
  18. 2025The Majority Vote Paradigm Shift: When Popular Meets Optimalcited by 0
  19. 2025Set a Thief to Catch a Thief: Combating Label Noise through Noisy Meta Learningcited by 0
  20. 2025Efficient Algorithms for Verifying Kruskal Rank in Sparse Linear Regression and Related Applicationscited by 0
  21. 2025Detect and Correct: A Selective Noise Correction Method for Learning with Noisy Labelscited by 0
  22. 2026SLT: Robust Quantum Neural Networks for Noisy-Label Medical Image Classification via Supermartingale-based Label Transitioncited by 0