Study reveals pervasive label errors in test sets, affecting machine learning benchmarks.
problem Label errors in test sets destabilize machine learning benchmarks.
method Identified label errors in 10 common datasets using confident learning algorithms and human validation.
result Lower capacity models may be more useful in real-world datasets with high proportions of erroneously labeled data.
Study shows label errors impact model disparity metrics, proposing mitigation methods.
problem Impact of label errors on model disparity metrics.
method Empirical study, characterizing label error effects; proposing estimation and relabeling methods.
result Label errors significantly affect model disparity metrics, particularly for minority groups.
This paper examines error bounds for deep learning classifiers with noisy labels.
problem Understanding the performance of classifiers trained on noisy data.
method Derives error bounds for excess risk, decomposing it into statistical and approximation errors. Uses independent block construction for statistical dependencies and vector-valued setting for approximation error.
result Established theoretical results for error bounds in deep learning with noisy labels, mitigating the impact of high-dimensional input spaces.
New algorithm controls type I error in NP classification under label noise.
problem Label noise affects NP classification methods, reducing power.
method Proposes a label-noise-adjusted Neyman-Pearson algorithm.
result Improves power while controlling type I error under desired level.
DP models misspecify LF dependencies, leading to significant performance errors.
problem Misspecification of LF dependencies in DP models.
method Theoretical bounds and empirical analysis of modeling errors.
result Modeling errors can be substantial, even with sensible LF structures.
Detects harmful shifts without labels for model performance.
problem Detecting distribution shifts without access to labels.
method Uses a proxy derived from predictions of an error estimator.
result High power and false alarm control under various shifts.
Estimates calibration error under label shift without labels.
problem Ensuring model reliability in the face of dataset shift without access to labels.
method Importance re-weighting of the labeled source distribution to estimate calibration error under label shift.
result Effective and reliable CE estimation with respect to the shifted target distribution.
ECN framework improves training on noisy structured labels.
problem Structured errors in fine-grained annotations lead to biased models.
method Error-Correcting Networks (ECN) framework.
result ECN improves fine-grained annotation prediction.
This paper examines how labeling error affects contrastive learning and proposes data dimensionality reduction methods to mitigate its impact.
problem The impact of labeling error on the performance of contrastive learning.
method Data dimensionality reduction methods (e.g., SVD) are applied to reduce false positive samples and improve downstream classification accuracy.
result Data dimensionality reduction methods can mitigate the negative impacts of labeling error on downstream classification performance.
Paper estimates optimal classification error with soft labels and calibration.
problem Estimating the optimal classification error with soft labels and calibration.
method Extends previous work on soft labels to estimate Bayes error, addressing bias and corrupted labels.
result The method provides a statistically consistent estimator of the Bayes error, even with imperfectly calibrated soft labels.
Develops a method for kernel ridge regression under covariate shift using pseudo-labels.
problem Learning a regression function with small mean squared error over a target distribution with labeled data from a different feature distribution.
method Split labeled data into two subsets, conduct kernel ridge regression on each, use imputation model to fill missing labels, and select the best candidate model.
result Non-asymptotic excess risk bounds demonstrate effective adaptation to target distribution and covariate shift.
Bayesian method improves segmentation accuracy with noisy labels.
problem Annotation errors in semantic segmentation due to mislabeling and spatial correlations.
method Approximate Bayesian estimation with spatially correlated discrete distributions and variational inference.
result The method achieves performance comparable to clean labels under moderate noise levels.
Learning exists in the context of data, yet notions of confidence typically focus on model predictions, not label quality. Confident learning (CL) is an alternative approach which focuses instead on label quality by characterizing and identifying label errors in datasets, based on the principles of pruning noisy data, …
Paper analyzes Gibbs and Langevin Monte Carlo for interpolation regime, showing generalization from low errors.
problem Analyzing Gibbs and Langevin Monte Carlo in overparameterized interpolation regime.
method Data-dependent bounds and stability under approximation with Langevin Monte Carlo.
result Generalization is signaled by small training errors in noisy regime, with bounds stable under approximation.
Semi-supervised learning improves with partial label information.
problem Improving model performance with limited labeled data.
method Contrastive learning with partial label information to encourage same labels.
result Partial label information reduces test error by up to 5.5 times.
New method improves false-/true-positive-rate estimation in fraud detection with noisy labels.
problem Estimating FPR/TPR in fraud detection with class-conditional label noise.
method Directly cleaning model's validation data to de-correlate cleaning error with model scores.
result Improves accuracy of FPR/TPR estimates, especially in asymmetric label noise scenarios.
Paper presents robust boosting methods for label noise.
problem Boosting methods degrade in noisy environments.
method Robust Minimax Boosting (RMBoost) with theoretical guarantees.
result RMBoost provides strong classification accuracy and robustness.
Study on noise models for noisy labels in NLP.
problem Quality of noise models from noisy labels.
method Theoretical analysis and synthetic dataset creation.
result Expected error of noise models derived.
SGD handles label noise with bounds improving over SGLD.
problem Label noise in non-convex optimization.
method Stochastic gradient descent with uniform dissipativity and smoothness conditions, using Wasserstein distance and algorithmic stability.
result Generalization error bounds with a rate of n−2/3, better than SGLD's n−1/2. CLSVAE repairs systematic errors in images with minimal labeled data.
problem Repairing systematic errors in data, especially in images.
method CLSVAE models inliers as a smaller latent space representation, separating inlier and outlier patterns.
result CLSVAE achieves superior repairs with less than 2% labeled data, outperforming other methods.
Paper bridges ordinary-label and complementary-label learning frameworks.
problem Combining complementary-label learning with ordinary-label learning.
method Integrates loss functions for one-versus-all and pairwise classification.
result Derives classification risk and error bound for additivity and duality loss functions.
A new metric predicts model performance on unseen data.
problem Predicting performance on out-of-distribution data without labels.
method Uses model predictions to pseudo-label data, trains a new model, and measures difference from in-distribution models.
result Empirically outperforms existing methods on image and text classification tasks.
Sharp bounds on uniform generalization errors in binary linear classification.
problem Understanding the uniform generalization errors in binary linear classification.
method Isoperimetric arguments, Poincaré and log-Sobolev inequalities for joint distributions.
result Sharp concentration bounds on uniform generalization errors, almost sure convergence in broad settings.
This paper refines human labeling as a measurement process, revealing four sources of variation.
problem Systematic variation in human labeling obscures model learning.
method Introduces a statistical framework to decompose labeling outcomes.
result Empirical evidence for four components of labeling variation.
Dual-T method improves transition matrix estimation in noisy label learning.
problem Large estimation error in noisy class posterior leads to poor transition matrix estimation.
method Introducing an intermediate class to avoid direct estimation of noisy class posterior, factorizing the transition matrix into two easier-to-estimate matrices.
result The dual-T estimator leads to better classification performances.
Unified view of label shift estimation methods.
problem Label distribution changes but class-conditional distributions remain the same.
method Unified view of two approaches: BBSE and MLLS.
result Unified framework and theoretical characterization of MLLS.
As sound event classification moves towards larger datasets, issues of label noise become inevitable. Web sites can supply large volumes of user-contributed audio and metadata, but inferring labels from this metadata introduces errors due to unreliable inputs, and limitations in the mapping. There is, however, little r…
New method selects data for labeling in RKHS to improve regression accuracy.
problem Labeling cost in supervised learning.
method Importance labeling scheme in RKHS with gradient descent.
result Gradient descent with proposed labeling scheme achieves optimal convergence rate.
This paper improves multi-label ranking by reweighting univariate losses, enhancing consistency and performance.
problem Improving multi-label ranking performance while maintaining consistency.
method Systematic study of consistency and generalization error bounds for learning algorithms, proposing a reweighted univariate loss.
result Inconsistent pairwise losses can lead to better performance than consistent univariate losses in practice.
We propose a streaming algorithm for the binary classification of data based on crowdsourcing. The algorithm learns the competence of each labeller by comparing her labels to those of other labellers on the same tasks and uses this information to minimize the prediction error rate on each task. We provide performance g…
Develops NPMC method for noisy labels, improving multiclass classification accuracy.
problem Asymmetric misclassification costs and label noise in multiclass classification.
method Empirical likelihood approach using exponential tilting density ratio model.
result Root n consistent and asymptotically normal estimators for clean labels and noise mechanism.
Binary classification improves with a small fraction of corrupted labels.
problem Binary classification with corrupted labels.
method Established corruption as a form of regularization and computed upper bounds on estimation error.
result Corruption is beneficial only up to a small fraction of the total sample, scaling with the square root of the sample size.
Improves label propagation for weakly supervised learning.
problem Reducing the need for labeled data in machine learning.
method Label Propagation with Weak Supervision (LPA) analysis.
result Demonstrated improvements over existing methods on weakly supervised classification tasks.
Non-convex SGD learns halfspaces with adversarial label noise efficiently.
problem Agnostically learning halfspaces in adversarial label noise settings.
method Non-convex SGD optimization for halfspace learning.
result Non-convex SGD achieves misclassification error close to optimal with adversarial noise.
Proposes a constrained labeling method for weakly supervised learning.
problem Combining weak supervision signals while navigating misleading correlations.
method Randomized constrained labeling within a defined space.
result Randomized constrained labeling converges after few iterations and outperforms other methods.
Deep Convolutional Neural Networks (CNN) enforces supervised information only at the output layer, and hidden layers are trained by back propagating the prediction error from the output layer without explicit supervision. We propose a supervised feature learning approach, Label Consistent Neural Network, which enforces…
Majority Vote is optimal for reliable data labeling under certain conditions.
problem Reliable data labeling requires aggregating multiple annotators' labels, but the optimality of Majority Vote is not well understood.
method Characterized conditions under which Majority Vote achieves the optimal label estimation error.
result Majority Vote optimally recovers labels for a given class distribution under tolerable annotation noise limits.
In this paper, we propose a new wrapper feature selection approach with partially labeled training examples where unlabeled observations are pseudo-labeled using the predictions of an initial classifier trained on the labeled training set. The wrapper is composed of a genetic algorithm for proposing new feature subsets…
Improved deep learning models with less labelled data and better label quality.
problem High costs and effort in training deep neural networks with label errors.
method Iterative label improvement using confidence-based filtering and dataset partitioning.
result Significant improvement in label quality and model accuracy.
This paper studies how label noise affects Federated Learning.
problem The impact of label noise on Federated Learning.
method The paper derives an upper bound for the generalization error and conducts experiments on MNIST and CIFAR-10 datasets.
result The global model accuracy decreases linearly with increasing label noise, consistent with theoretical analysis.
Efficiently tests two distributions with few label queries.
problem Two-sample test with limited label information.
method Three-stage framework: classifier training, bimodal query, FR test.
result Significantly reduces Type II error compared to uniform querying.
Semi-supervised learning methods are motivated by the availability of large datasets with unlabeled features in addition to labeled data. Unlabeled data is, however, not guaranteed to improve classification performance and has in fact been reported to impair the performance in certain cases. A fundamental source of err…
While crowdsourcing has become an important means to label data, there is great interest in estimating the ground truth from unreliable labels produced by crowdworkers. The Dawid and Skene (DS) model is one of the most well-known models in the study of crowdsourcing. Despite its practical popularity, theoretical error …
FLOPART solves peak detection by creating accurate train and test set predictions.
problem Correctly detecting peaks in sequential data.
method Dynamic programming changepoint algorithm with zero train label errors.
result FLOPART provides highly accurate predictions on both train and test sets.
Paper tackles active labeling for partial supervision.
problem Accessing stochastic gradients with partial supervision.
method Streaming technique to minimize generalization error.
result Proves minimization of generalization error ratio.
Study shows SQ hardness for multiclass linear classification with random noise.
problem Complexity of multiclass linear classification with random noise.
method Proves super-polynomial SQ lower bounds for MLC with RCN.
result Super-polynomial SQ lower bounds for MLC with RCN.
Paper improves MRI reconstruction by separating target labels and prediction error.
problem Improving MRI reconstruction accuracy by estimating prediction error.
method Proposes a novel method to estimate target labels and prediction error separately.
result Significantly better MRI reconstruction results achieved compared to state-of-the-art methods.
Calibration without labels in multiple testing
problem Interpretable error probabilities in large-scale hypothesis testing
method Constructing pseudo-labels from spacings of ordered p-values result Finding that q-value can be severely miscalibrated