arXiv research

This paper establishes lower bounds for SGD's error, matching upper bounds.

Paper establishes universal lower bounds and optimal rates for clustering sub-exponential mixture models.

Study controls error rates of binary classifiers using hypothesis testing.

We carefully study how well minimizing convex surrogate loss functions, corresponds to minimizing the misclassification error rate for the problem of binary classification with linear predictors. In particular, we show that amongst all convex surrogate losses, the hinge loss gives essentially the best possible bound, o…

Optimal number of voters for a voting ensemble can be estimated from the distribution of classifier errors.

This paper tackles worst-class error rate in classification tasks.

The paper bounds neural networks' approximation error and applies it to regression and GANs.

Quantum codes on hyperbolic lattices outperform Euclidean ones with higher rates and lower overhead.

Paper establishes a universal growth rate for smooth surrogate losses in classification.

New algorithm optimally identifies best arm in both stochastic and adversarial settings.

We show that if $F$ is a convex class of functions that is $L$-subgaussian, the error rate of learning problems generated by independent noise is equivalent to a fixed point determined by `local' covering estimates of the class, rather than by the gaussian averages. To that end, we establish new sharp upper and lower e…

Improves bit error tolerance in RRAM-based BNNs without overfitting.

Deep convolutional architecture identifies eye movements for biometric faster and more accurately.

New evidence shows computational barriers in graphon estimation using low-degree polynomials.

In this paper, we consider a popular model for collaborative filtering in recommender systems where some users of a website rate some items, such as movies, and the goal is to recover the ratings of some or all of the unrated items of each user. In particular, we consider both the clustering model, where only users (or…

Nonasymptotic error bounds and strong consistency rates for survival analysis methods.

Paper analyzes minimax error rate for crowdsourcing models.

Lower bounds on Bayes risk for realizable models derived using information theory.

GANs learn distributions well from samples, with rates depending on intrinsic dimension.

Paper develops a framework to derive lower bounds on FDR and FNR in multiple testing.

Paper proposes deep neural networks for nonparametric regression from dependent data.

Ensembling improves performance when classifiers disagree more than average.

In this paper, we initiate a systematic investigation of differentially private algorithms for convex empirical risk minimization. Various instantiations of this problem have been studied before. We provide new algorithms and matching lower bounds for private ERM assuming only that each data point's contribution to the…

For the problem of high-dimensional sparse linear regression, it is known that an $\ell_0$-based estimator can achieve a $1/n$ "fast" rate on the prediction error without any conditions on the design matrix, whereas in absence of restrictive conditions on the design matrix, popular polynomial-time methods only guarante…

The study examines conditions for achieving a simple lower bound in estimating mean from samples.

Crowdsourcing has become an effective and popular tool for human-powered computation to label large datasets. Since the workers can be unreliable, it is common in crowdsourcing to assign multiple workers to one task, and to aggregate the labels in order to obtain results of high quality. In this paper, we provide finit…

Corrected samplers reduce discretization error in discrete flow models without additional computational cost.

The paper analyzes kNN density estimation's convergence rates under different conditions.

Simplifies VAE for anomaly detection using rate-distortion theory.

Kernel interpolation is inconsistent for norms with smoothness above a constant.

The paper characterizes a fundamental tradeoff between fairness and accuracy in classification problems.

In many machine learning applications, crowdsourcing has become the primary means for label collection. In this paper, we study the optimal error rate for aggregating labels provided by a set of non-expert workers. Under the classic Dawid-Skene model, we establish matching upper and lower bounds with an exact exponent …

Active data collection improves convergence rates in operator learning.

This study analyzes how well GANs approximate distributions from small samples.

Paper analyzes kNN estimator for KL divergence, proving its optimality.

Study improves weak error estimates for rough volatility models.

Consider the problem of pricing options on forwards in energy markets, when spot prices follow a geometric multi-factor model in which several rates of mean reversion appear. In this paper we investigate the role played by slow mean reversion when pricing and hedging options. In particular, we determine both upper and …

Semi-supervised learning can't outperform supervised learning for 2-Gaussian mixture models.

Hallucinations in models are mislinked estimates, not errors.

SGD with random shuffling achieves lower optimization error than repeated shuffling.

New findings on complexity limits in fixed budget bandit identification.

MACI improves LLM factuality inference with higher retention and lower time cost.

SDP achieves Bayes error rate in synchronization and block models.

This paper closes the gap on matching pursuit's convergence rate.

Improved efficient robust regression with near-linear time and subquadratic samples.

This paper examines fundamental error characteristics for a general class of matrix completion problems, where the matrix of interest is a product of two a priori unknown matrices, one of which is sparse, and the observations are noisy. Our main contributions come in the form of minimax lower bounds for the expected pe…

New algorithm improves regression error bounds and accelerates performance for low noise.

The study establishes minimax bounds for estimating operators from noisy samples.