arXiv research

The paper cleans label noise in supervised classification using Bernoulli sampling.

Hierarchical randomized smoothing improves model robustness for complex data.

NoiseRank reduces label noise without supervision, improving classification accuracy.

Study shows a tradeoff between sample complexity and computational efficiency for learning halfspaces with random noise.

Study near-optimal bounds for learning Gaussian halfspaces with random noise.

Study shows SQ hardness for multiclass linear classification with random noise.

We describe a framework for designing efficient active learning algorithms that are tolerant to random classification noise and are differentially-private. The framework is based on active learning algorithms that are statistical in the sense that they rely on estimates of expectations of functions of filtered random e…

SGLRW improves robustness of stochastic gradient MCMC methods.

Although kernel methods are widely used in many learning problems, they have poor scalability to large datasets. To address this problem, sketching and stochastic gradient methods are the most commonly used techniques to derive efficient large-scale learning algorithms. In this study, we consider solving a binary class…

Unhinged loss minimization fails to improve classifier accuracy for simple data.

Several recent works have shown that state-of-the-art classifiers are vulnerable to worst-case (i.e., adversarial) perturbations of the datapoints. On the other hand, it has been empirically observed that these same classifiers are relatively robust to random noise. In this paper, we propose to study a \textit{semi-ran…

This work introduces a noise-adaptive conformal inference method for better prediction sets in noisy data.

This paper studies the classification of high-dimensional Gaussian signals from low-dimensional noisy, linear measurements. In particular, it provides upper bounds (sufficient conditions) on the number of measurements required to drive the probability of misclassification to zero in the low-noise regime, both for rando…

Recently, techniques have been developed to provably guarantee the robustness of a classifier to adversarial perturbations of bounded L_1 and L_2 magnitudes by using randomized smoothing: the robust classification is a consensus of base classifications on randomly noised samples where the noise is additive. In this pap…

We introduce a training method for both better word representation and performance, which we call GROVER (Gradual Rumination On the Vector with maskERs). The method is to gradually and iteratively add random noises to word embeddings while training a model. GROVER first starts from conventional training process, and th…

Model classifies hand gestures with low EMG samples augmented with noise.

In this paper, we study a classification problem in which sample labels are randomly corrupted. In this scenario, there is an unobservable sample with noise-free labels. However, before being observed, the true labels are independently flipped with a probability $ρ\in[0,0.5)$, and the random label noise can be class-co…

New algorithm reduces feature count and accelerates error convergence.

Study shows label noise impacts neural representations' information content, revealing double descent behavior.

Study examines noise sensitivity of DNNs for binary classification.

New algorithm uses random projections for robust, sparse data classification.

The paper extends Pearson correlation to multi-variables, useful for noise measurement and feature selection.

The paper establishes risk bounds for PU learning with label noise.

The paper proposes a method to detect and filter noisy or mislabeled data using pointwise mutual information.

Logistic regression can handle noisy labels effectively when labels are imperfectly assigned by multiple experts.

DiPriMe forests use private medians to create balanced tree splits for privacy-protected data.

CNT leverages noisy targets to guide model learning.

Spectral clustering performance depends on eigenvector fluctuations, shown to be Gaussian.

Random Fourier features classification achieves fast learning rates with fewer features.

CNNs learn about input data uncertainties, improving classification performance.

We show how to turn any classifier that classifies well under Gaussian noise into a new classifier that is certifiably robust to adversarial perturbations under the $\ell_2$ norm. This "randomized smoothing" technique has been proposed recently in the literature, but existing guarantees are loose. We prove a tight robu…

Study compares classification techniques to predict customer churn in banking.

Randomized classifiers improve strategic classification efficiency.

This paper connects noise injection to Bayesian inference for neural networks, improving model uncertainty.

Rocket algorithm classifies time-series data efficiently using random projections and natural sparsity.

Image classification systems recently made a giant leap with the advancement of deep neural networks. However, these systems require an excessive amount of labeled data to be adequately trained. Gathering a correctly annotated dataset is not always feasible due to several factors, such as the expensiveness of the label…

Method matches noisy remote sensing images robustly.

Dual-T method improves transition matrix estimation in noisy label learning.

Adaptive denoising models adjust the number of steps based on noise level.

GraphMoE generates random graphs using neural networks and graphlets.

Boosting algorithm reduces error in noisy data.

We introduce a simple framework for designing private boosting algorithms. We give natural conditions under which these algorithms are differentially private, efficient, and noise-tolerant PAC learners. To demonstrate our framework, we use it to construct noise-tolerant and private PAC learners for large-margin halfspa…

Existence of strong randomized equilibria in mean-field games with common noise.

New method enhances neural network robustness against adversarial attacks.

DP-RandP improves privacy-utility tradeoff in DP-SGD by learning priors from random processes.

New methods for handling time-varying label noise in time series classification.

RS-NSGD improves SGD convergence for heavy-tailed noise.

Many modern machine learning models are trained to achieve zero or near-zero training error in order to obtain near-optimal (but non-zero) test error. This phenomenon of strong generalization performance for "overfitted" / interpolated classifiers appears to be ubiquitous in high-dimensional data, having been observed …