arXiv research

For binary classification we establish learning rates up to the order of $n^{-1}$ for support vector machines (SVMs) with hinge loss and Gaussian RBF kernels. These rates are in terms of two assumptions on the considered distributions: Tsybakov's noise assumption to establish a small estimation error, and a new geometr…

We study Principal Component Analysis (PCA) in a setting where a part of the corrupting noise is data-dependent and, as a result, the noise and the true data are correlated. Under a bounded-ness assumption on the true data and the noise, and a simple assumption on data-noise correlation, we obtain a nearly optimal samp…

Enhanced consistency bounds derived for classification under a new noise condition.

New method bounds hardware noise without assumptions.

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

Study examines robustness of NPI effectiveness models against COVID-19.

Compression is at the heart of effective representation learning. However, lossy compression is typically achieved through simple parametric models like Gaussian noise to preserve analytic tractability, and the limitations this imposes on learning are largely unexplored. Further, the Gaussian prior assumptions in model…

New findings show optimal noise in contrastive learning is not the same as data distribution.

New algorithms SVCA and SSPA improve robustness to noise in nonnegative matrix factorization.

Unified framework for isotropic SG noise in posterior sampling.

New approach uses 'forward-looking' counterfactuals for treatment choice.

New bandit algorithm for non-i.i.d. noise, improving standard rates.

Boosting algorithm reduces error in noisy data.

Paper studies Adam's convergence under relaxed assumptions, proving a rate of O(poly(log T)/sqrt(T)).

We consider classification in the presence of class-dependent asymmetric label noise with unknown noise probabilities. In this setting, identifiability conditions are known, but additional assumptions were shown to be required for finite sample rates, and so far only the parametric rate has been obtained. Assuming thes…

Study revisits AdaGrad convergence with relaxed noise assumptions.

Noise makes learning linear thresholds hard, but algorithms can still learn near-optimal thresholds.

Calibration of simplified vine copulas using noise contrastive estimation

In high-dimensional data, structured noise caused by observed and unobserved factors affecting multiple target variables simultaneously, imposes a serious challenge for modeling, by masking the often weak signal. Therefore, (1) explaining away the structured noise in multiple-output regression is of paramount importanc…

Study on ReLU regression with Massart noise, achieving exact parameter recovery.

Improves generative models by using heavy-tailed noise in score matching.

New method reduces variance in stochastic optimization with high confidence.

Classical scaling is shown to be optimal under various noisy conditions.

New research shows that binary classification can be done with noisy data, but only if there are clean samples available.

Researchers establish bounds for SGMs' KL and Wasserstein divergences under various noise schedules.

Free lunch from noise reveals linear spectral features for RL.

Bayes-optimal limits in PCA with structured noise are determined.

BDDMs eliminate noise conditioning in diffusion models, simplifying training and sampling.

New method learns DAGs from noisy data without identifiability assumptions.

We consider a high dimensional linear regression problem where the goal is to efficiently recover an unknown vector $β^*$ from $n$ noisy linear observations $Y=Xβ^*+W \in \mathbb{R}^n$, for known $X \in \mathbb{R}^{n \times p}$ and unknown $W \in \mathbb{R}^n$. Unlike most of the literature on this model we make no spa…

Last-iterate guarantees for learning in co-coercive games under noisy feedback.

Study improves robustness and sparsity in linear regression with adversarial outliers and heavy-tailed noise.

Stochastic gradient methods can converge in expectation under heavy-tailed noise.

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

New method improves multi-fidelity Bayesian optimization by accounting for local correlations and varying noise.

Paper analyzes convergence of stochastic methods under heavy-tailed noise.

Fairness-aware learning involves designing algorithms that do not discriminate with respect to some sensitive feature (e.g., race or gender). Existing work on the problem operates under the assumption that the sensitive feature available in one's training sample is perfectly reliable. This assumption may be violated in…

Interpolating label noise makes models vulnerable to adversarial attacks.

Principal Component Analysis (PCA) has wide applications in machine learning, text mining and computer vision. Classical PCA based on a Gaussian noise model is fragile to noise of large magnitude. Laplace noise assumption based PCA methods cannot deal with dense noise effectively. In this paper, we propose Cauchy Princ…

Introduces CCR for constructing confidence regions from conformal predictions.

Improves fault detection models in noisy data.

Develops efficient inference for noise heterogeneity in machine learning models.

Paper reconciles minimax rates and optimal recovery rates for noisy observations.

ALPCAH improves PCA for noisy data by estimating sample-wise noise variances.

In learning with noisy labels, for every instance, its label can randomly walk to other classes following a transition distribution which is named a noise model. Well-studied noise models are all instance-independent, namely, the transition depends only on the original label but not the instance itself, and thus they a…

Study large deviations rates for SGD with strongly convex functions.

Efficiently estimates sparse linear regression with heavy-tailed and outlier-contaminated data.

Estimates shared linear subspace from noisy data with multiple users.