mlsanityData to 7 Oct 2026

On Symmetric Losses for Learning from Corrupted Labels

arXiv 1901.09314 ↗First posted 27 Jan 2019stat.ML cs.LGcited by 53

Not on the map. Why?

Abstract (arXiv)

This paper aims to provide a better understanding of a symmetric loss. First, we emphasize that using a symmetric loss is advantageous in the balanced error rate (BER) minimization and area under the receiver operating characteristic curve (AUC) maximization from corrupted labels. Second, we prove general theoretical properties of symmetric losses, including a classification-calibration condition, excess risk bound, conditional risk minimizer, and AUC-consistency condition. Third, since all nonnegative symmetric losses are non-convex, we propose a convex barrier hinge loss that benefits significantly from the symmetric condition, although it is not symmetric everywhere. Finally, we conduct experiments to validate the relevance of the symmetric condition.

Key facts. Cited by 53 papers in the map (data through 7 Oct 2026). First posted on arXiv 27 Jan 2019.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2020: 1; 2021: 15; 2022: 12; 2023: 11; 2024: 3; 2025: 5; 2026: 6.

53 papers.

Sort:

By area:

  1. 2021Learning with Feature-Dependent Label Noise: A Progressive Approachcited by 102
  2. 2022AUC Maximization in the Era of Big Data and AI: A Surveycited by 100
  3. 2021Learning Noise Transition Matrix from Only Noisy Labels via Total Variation Regularizationcited by 62
  4. 2021Asymmetric Loss Functions for Learning with Noisy Labelscited by 46
  5. 2021Learning with Noisy Labels via Sparse Regularizationcited by 38
  6. 2020Robust Imitation Learning from Noisy Demonstrationscited by 26
  7. 2021Understanding Instance-Level Label Noise: Disparate Impacts and Treatmentscited by 22
  8. 2021Adaptive Sample Selection for Robust Learning under Label Noisecited by 20
  9. 2022Boosting Facial Expression Recognition by A Semi-Supervised Progressive Teachercited by 12
  10. 2023Weakly Supervised AUC Optimization: A Unified Partial AUC Approachcited by 11
  11. 2022Benchmarking Deep AUROC Optimization: Loss Functions and Algorithmic Choicescited by 10
  12. 2022Learning from Multiple Annotator Noisy Labels via Sample-wise Label Fusioncited by 10
38 more
  1. 2022FeDXL: Provable Federated Learning for Deep X-Risk Optimizationcited by 10
  2. 2021Robustness and Reliability When Training With Noisy Labelscited by 9
  3. 2021Learning with Label Noise for Image Retrieval by Selecting Interactionscited by 8
  4. 2021Learning with Proper Partial Labelscited by 8
  5. 2023NoisywikiHow: A Benchmark for Learning with Real-world Noisy Labels in Natural Language Processingcited by 8
  6. 2023Regularly Truncated M-estimators for Learning with Noisy Labelscited by 8
  7. 2022Class-Imbalanced Complementary-Label Learning via Weighted Losscited by 7
  8. 2023Confidence-based Reliable Learning under Dual Noisescited by 6
  9. 2025Frequency-Semantic Enhanced Variational Autoencoder for Zero-Shot Skeleton-based Action Recognitioncited by 6
  10. 2023Unified Risk Analysis for Weakly Supervised Learningcited by 5
  11. 2022Prototype-Anchored Learning for Learning with Imperfect Annotationscited by 4
  12. 2023SANTA: Separate Strategies for Inaccurate and Incomplete Annotation Noise in Distantly-Supervised Named Entity Recognitioncited by 4
  13. 2023Making Binary Classification from Multiple Unlabeled Datasets Almost Free of Supervisioncited by 4
  14. 2025On Symmetric Losses for Robust Policy Optimization with Noisy Preferencescited by 4
  15. 2021Approximating Instance-Dependent Noise via Instance-Confidence Embeddingcited by 3
  16. 2021Do We Really Need Gold Samples for Sample Weighting Under Label Noise?cited by 3
  17. 2023Provable Multi-instance Deep AUC Maximization with Stochastic Poolingcited by 3
  18. 2024On the use of Silver Standard Data for Zero-shot Classification Tasks in Information Extractioncited by 3
  19. 2024Active Negative Loss: A Robust Framework for Learning with Noisy Labelscited by 3
  20. 2021An Exploration into why Output Regularization Mitigates Label Noisecited by 2
  21. 2021The perils of being unhinged: On the accuracy of classifiers minimizing a noise-robust convex losscited by 2
  22. 2025Establishing Linear Surrogate Regret Bounds for Convex Smooth Losses via Convolutional Fenchel-Young Lossescited by 2
  23. 2022What killed the Convex Booster ?cited by 1
  24. 2022Maximising the Utility of Validation Sets for Imbalanced Noisy-label Meta-learningcited by 1
  25. 2024Exploring Loss Design Techniques For Decision Tree Robustness To Label Noisecited by 1
  26. 2025Accessible, Realistic, and Fair Evaluation of Positive-Unlabeled Learning Algorithmscited by 1
  27. 2021A Symmetric Loss Perspective of Reliable Machine Learningcited by 0
  28. 2021Contrastive Representations for Label Noise Require Fine-Tuningcited by 0
  29. 2022Deep Learning with Label Noise: A Hierarchical Approachcited by 0
  30. 2022Performances of Symmetric Loss for Private Data from Exponential Mechanismcited by 0
  31. 2022Learning with Silver Standard Data for Zero-shot Relation Extractioncited by 0
  32. 2023Learning from Stochastic Labelscited by 0
  33. 2023AUC Optimization from Multiple Unlabeled Datasetscited by 0
  34. 2023Domain Transfer Through Image-to-Image Translation for Uncertainty-Aware Prostate Cancer Classificationcited by 0
  35. 2025Learning from Uncertain Similarity and Unlabeled Datacited by 0
  36. 2026Communication-Efficient Federated AUC Maximization with Cyclic Client Participationcited by 0
  37. 2026Generalized Distribution-Free Semi-Supervised Learning with Risk Rewritecited by 0
  38. 2026Robust Losses from Univariate Base Functions for Noisy-Label Learningcited by 0

50 most cited of 53 All 53 (JSON)