Research
On-device research index

arXiv research

A locally-built, LLM-digested index of recent arXiv papers in quant finance, geometry/topology, and statistical ML — keyword search served straight from SQLite on this machine.

169,181 papers · 148 categories

Trend · papers per month

225450674899 · Jun 202019922001200920182026
48 results for noise training

Noise improves model quality in non-linear neural networks during decentralized training.

problem Improving generalization of locally trained neural networks.
method Injecting noise into the weights of neural networks during decentralized training.
result Noise injection improves model quality for non-linear neural networks, but not for linear models.

Paper proposes a method to train deep text classification models robust to label noise.

problem Training deep text classification models with noisy labels.
method Introduces a non-linear processing layer (noise model) into CNN architecture, learned jointly with CNN weights.
result The approach enables better sentence representations and robustness to extreme label noise.

Synthetic noise training improves machine translation robustness to spelling mistakes.

problem Making machine translation robust to spelling mistakes and natural noise.
method Training on synthetic noise to improve robustness to natural noise.
result Training on synthetic noise improves robustness to natural noise without diminishing performance on clean text.

Improved neural language models trained with dynamic noise-contrastive estimation.

problem Training large-scale language models efficiently and avoiding overfitting.
method Dynamic Noise-Contrastive Estimation (DNCE) to train neural trans-dimensional random field language models.
result DNCE reduces training cost and improves model performance on large datasets.

FANNet analyzes noise tolerance and training bias in neural networks.

problem Low noise tolerance and input sensitivity in neural networks lead to failures on unseen inputs.
method Formal analysis using model checking under different noise ranges.
result Noise tolerance of ±11%\pm 11\% for the trained network, sensitive input nodes identified, and biasness confirmed.

TrustNet robustly learns noise patterns from trusted data to improve weakly-supervised classification.

problem Robustness to label noise in weakly-supervised learning.
method TrustNet learns noise patterns from trusted data, then trains a robust classifier using these patterns.
result TrustNet outperforms state-of-the-art methods in robustness to various noise patterns.

S-SGD adds symmetrical noise to weights to avoid sharp minima in deep learning.

problem SGD does not always converge to a flat minimum, leading to poor generalization.
method Symmetrical weight noise injection in SGD.
result S-SGD outperforms conventional SGD and weight-noise injection methods in large batch training.

Analytic expressions for deep neural network output under stochastic training.

problem Understanding the impact of noise and hyperparameters on deep neural network performance.
method Taylor expansion of network output to derive analytical expressions for weights and output.
result Noise in training affects generalization by preventing the output from fully converging on train data, but does not provide explicit regularization.

New bounds for KANs trained with DP-SGD, addressing correlated noise.

problem Risk bounds for Kolmogorov-Arnold Networks trained by DP-SGD with correlated noise.
method Established new optimization and population risk analysis for KANs trained with DP-SGD, addressing correlated noise.
result First optimization and population risk analysis of correlated-noise mechanisms for DP training in non-convex settings, including neural networks.

RAD improves robustness to domain annotation noise without explicit domain annotations.

problem Robustness to domain annotation noise in training data.
method Regularized Annotation of Domains (RAD) for last layer retraining.
result RAD outperforms state-of-the-art methods even with 5% noise in training data.

Interpolating noisy data in linear regression leads to zero training error.

problem Understanding why deep neural networks generalize well with noisy data.
method Investigated overparameterized linear regression, analyzing generalization error and proposing a hybrid scheme.
result Interpolating solutions in noisy data can generalize well, with error decaying to zero with more features.

Early stopping helps prevent overfitting to noisy labels in neural networks.

problem Overfitting to noisy labels in real-world training data.
method Two-phase training method (Prestopping) that early stops training and resumes using a maximal safe set.
result Significantly outperforms state-of-the-art methods in test error under label noise.

The study improves generalization in large-batch training by adding structured covariance noise to gradients.

problem Improving generalization in large-batch training while maintaining optimal convergence.
method Adding covariance noise to the gradients to improve generalization performance.
result The method improves generalization performance without degrading optimization performance and training duration.

Large datasets often have unreliable labels-such as those obtained from Amazon's Mechanical Turk or social media platforms-and classifiers trained on mislabeled datasets often exhibit poor performance. We present a simple, effective technique for accounting for label noise when training deep neural networks. We augment…

2017-05-09abs ↗pdf ↗

Paper tackles robustness in adversarial noise with a meta-optimizer.

problem Sensitivity to adversarial noise hinders machine learning deployment.
method Meta-optimizer learns to robustly optimize models using adversarial examples.
result Meta-optimizer transfers adversarial knowledge to new models without generating new examples.

SGD-trained neural networks generalize well even with adversarial label noise.

problem Generalization of neural networks trained on adversarial label noise.
method Training a one-hidden-layer neural network with SGD on arbitrary width networks.
result SGD-trained networks achieve classification accuracy competitive with the best halfspace over adversarial label noise.

IEBN normalizes noise by enhancing instance-specific information, improving deep learning performance.

problem Improving deep learning performance by regulating noise in batch normalization.
method Integrates self-attention mechanism to recalibrate channel information in BN.
result IEBN outperforms BN with improved generalization and stability.

Stochastic Gradient Descent introduces noise in training, affecting model decision boundaries.

problem Understanding the impact of noise in SGD on model decision boundaries.
method Characterized SGD and persistent SGD dynamics in a neural network model, measuring noise magnitude in both under- and over-parametrized regimes.
result Noisier algorithms lead to wider decision boundaries in constraint satisfaction problems.

Extends neural network training framework to handle noise and uncertainty.

problem Handling noise and uncertainty in neural network training.
method Integrates non-zero aleatoric noise and derives posterior covariance for epistemic uncertainty.
result Derives an estimator for posterior covariance, providing a handle on epistemic uncertainty.

Self-supervised method predicts clean signal and noise distribution from noisy images.

problem Blind denoising and noise estimation in biomedical images with limited clean data.
method Two neural networks jointly predict clean signal and noise distribution from noisy observations.
result Significantly outperforms state-of-the-art algorithms on six biomedical image datasets.

This work improves ASR noise robustness using parallel data and T/S learning.

problem Noise robustness in automatic speech recognition.
method Teacher-student learning with parallel clean and noisy data, logits selection.
result Best student model yields significant WER reductions in noisy conditions.

Noise improves deep neural network performance, especially in knowledge distillation.

problem Improving deep neural network performance and reducing performance gap.
method Injecting constructive noise at different levels in the collaborative learning framework.
result Constructive noise enables effective training and distillation of desirable characteristics.

NoL approach improves adversarial robustness by modeling random noise during training.

problem Improving neural network robustness against adversarial attacks.
method Implicit generative modeling of random noise during training.
result Models trained with NoL perform better against a wide range of adversarial attacks.

CNNs trained by gradient descent can learn intrinsic image rank robustly to background noises.

problem Understanding the intrinsic dimension of data in over-parameterized CNNs.
method Theoretical analysis and experiments on synthetic and real datasets.
result CNNs trained by gradient descent can learn the intrinsic dimension of clean images robustly to background noises.

New method calibrates noise for attack risk, improving ML model accuracy.

problem Improving accuracy of privacy-preserving ML models while maintaining privacy.
method Directly calibrates noise scale to a desired attack risk level, bypassing the standard ε\varepsilon-calibration.
result Significantly decreases noise scale, leading to increased utility at the same risk level.

Paper improves image classification accuracy with a new Noise Modeling Network.

problem Improving performance of multi-label image classifiers with noisy or missing labels.
method Integrates a Noise Modeling Network (NMN) with a CNN to jointly learn noise distribution and CNN parameters.
result Consistently improves classification performance on MSR-COCO and MSR-VTT datasets.

Colored noise improves neural network robustness against adversarial attacks.

problem Vulnerability of neural networks to adversarial perturbations.
method Injection of colored noise into network weights and activations during adversarial training.
result Our approach outperforms previous methods in terms of adversarial accuracy on CIFAR-10 and CIFAR-100 datasets.

DynaCor detects noisy labels by learning from corrupted training signals.

problem Label noise in real-world datasets hinders model generalization.
method DynaCor introduces label corruption to indirectly simulate noisy labels and learns to distinguish clean from noisy instances.
result DynaCor outperforms state-of-the-art competitors in noisy label detection.