Abstract:Label noise is ubiquitous in many real-world scenarios which often misleads training algorithm and brings about the degraded classification performance. Therefore, many approaches have been proposed to correct the loss function given corrupted labels to combat such label noise. Among them, a trend of works achieve this goal by unbiasedly estimating the data centroid, which plays an important role in constructing an unbiased risk estimator for minimization. However, they usually handle the noisy labels in different classes all at once, so the local information inherited by each class is ignored which often leads to unsatisfactory performance. To address this defect, this paper presents a novel robust learning algorithm dubbed "Class-Wise Denoising" (CWD), which tackles the noisy labels in a class-wise way to ease the entire noise correction task. Specifically, two virtual auxiliary sets are respectively constructed by presuming that the positive and negative labels in the training set are clean, so the original false-negative labels and false-positive ones are tackled separately. As a result, an improved centroid estimator can be designed which helps to yield more accurate risk estimator. Theoretically, we prove that: 1) the variance in centroid estimation can often be reduced by our CWD when compared with existing methods with unbiased centroid estimator; and 2) the performance of CWD trained on the noisy set will converge to that of the optimal classifier trained on the clean set with a convergence rate O(1/vn) )where n is the number of the training examples. These sound theoretical properties critically enable our CWD to produce the improved classification performance under label noise, which is also demonstrated by the comparisons with ten representative state-of-the-art methods on a variety of benchmark datasets.

Online Multi-Label Classification under Noisy and Changing Label Distribution

Holistic Label Correction for Noisy Multi-Label Classification

Online Passive-Aggressive Multilabel Classification Algorithms

Adaptive label thresholding methods for online multi-label classification.

Enhancing Multi-Label Text Classification under Label-Dependent Noise: A Label-Specific Denoising Framework

Imbalanced Multiple Noisy Labeling

Towards Robust Multi-Label Learning Against Dirty Label Noise

Handling New Class in Online Label Shift

Imbalanced Multiple Noisy Labeling for Supervised Learning.

Multi-task Label Noise Learning for Classification

Scalable Label Distribution Learning for Multi-Label Classification

Class-Wise Denoising for Robust Learning under Label Noise

Multi-class Label Noise Learning via Loss Decomposition and Centroid Estimation

Hybrid Noise-Oriented Multilabel Learning.

Online Semi-Supervised Multi-Label Classification with Label Compression and Local Smooth Regression

Active Learning with Imbalanced Multiple Noisy Labeling

Fine-Grained Classification with Noisy Labels

Two-stage Natural Scene Image Classification with Noise Discovering and Label-Correlation Mining

Neighborhood Collective Estimation for Noisy Label Identification and Correction

Feature-Induced Label Distribution for Learning with Noisy Labels

Adapting to Online Label Shift with Provable Guarantees