arXiv research

Proves subgaussian distributions are SoS-certifiably subgaussian, enabling efficient algorithms for various statistical tasks.

Thompson Sampling shows polynomial regret for combinatorial semi-bandits with subgaussian rewards.

Robustly estimates linear regression coefficients with adversarial and noisy data.

In this note, we derive concentration inequalities for random vectors with subGaussian norm (a generalization of both subGaussian random vectors and norm bounded random vectors), which are tight up to logarithmic factors.

Improved mean estimation for symmetric distributions with finite-sample guarantees.

Note on subgaussian bounds for sign-quantized linear maps.

Ridge regression performs optimally in noisy environments with heavy-tailed distributions.

We study tilting subweibull distributions and their tail behavior.

New robust estimators achieve subgaussian bounds using VC-dimension.

Bandit algorithms struggle with consistent performance and robustness.

New bounds derived for machine learning algorithms using convex functions.

We solve ReLU regression with efficient approximations for various distributions.

Estimates change point in high dimensional time series models.

New algorithm for robust regression with subgaussian error bound.

New algorithms allocate sampling budget to estimate group means without exploration.

New algorithms robustly estimate mean with near-optimal error rates.

Deep neural networks help recover two signals from noisy mixtures.

Paper estimates EOT maps for non-compactly supported measures with subGaussian target.

Diffusion models learn multi-modal distributions with optimal efficiency.

Estimates score function from data with optimal rate in high dimensions.

This work establishes near-minimax optimal guarantees for ODE-based samplers under mild assumptions.

Many theoretical results on estimation of high dimensional time series require specifying an underlying data generating model (DGM). Instead, along the footsteps of~\cite{wong2017lasso}, this paper relies only on (strict) stationarity and $β$-mixing condition to establish consistency of lasso when data comes from a $β…

We present a theory for Euclidean dimensionality reduction with subgaussian matrices which unifies several restricted isometry property and Johnson-Lindenstrauss type results obtained earlier for specific data sets. In particular, we recover and, in several cases, improve results for sets of sparse and structured spars…

The new field of adaptive data analysis seeks to provide algorithms and provable guarantees for models of machine learning that allow researchers to reuse their data, which normally falls outside of the usual statistical paradigm of static data analysis. In 2014, Dwork, Feldman, Hardt, Pitassi, Reingold and Roth introd…

New algorithms exploit mean bounds to improve bandit problem performance.

This paper tackles open problem of tight bounds for KBs with Bernoulli rewards.

We introduce a model-free relax-and-round algorithm for k-means clustering based on a semidefinite relaxation due to Peng and Wei. The algorithm interprets the SDP output as a denoised version of the original data and then rounds this output to a hard clustering. We provide a generic method for proving performance guar…

Estimates mean from heavy-tailed data without variance.

RONM method reduces regret in stochastic convex bandits with decreasing noise.

The paper provides entrywise bounds for Sparse PCA, improving upon previous results.

Polynomial-time private algorithm for robust estimation of mean and covariance in the presence of outliers.

New SQ lower bound shows complexity nearly matches known upper bound for smoothed agnostic learning.

Paper designs a bandit algorithm without reward distribution info.

New algorithm optimizes convex functions with noisy evaluations in one dimension.

This work studies applications and generalizations of a simple estimation technique that provides exponential concentration under heavy-tailed distributions, assuming only bounded low-order moments. We show that the technique can be used for approximate minimization of smooth and strongly convex losses, and specificall…

Existing strategies for finite-armed stochastic bandits mostly depend on a parameter of scale that must be known in advance. Sometimes this is in the form of a bound on the payoffs, or the knowledge of a variance or subgaussian parameter. The notable exceptions are the analysis of Gaussian bandits with unknown mean and…

New coherence parameter for GNNs with Fourier measurements improves signal recovery.

Study finds the minimum number of finite Gaussian mixtures for best approximation.

This note gives a simple analysis of a randomized approximation scheme for matrix multiplication proposed by Sarlos (2006) based on a random rotation followed by uniform column sampling. The result follows from a matrix version of Bernstein's inequality and a tail inequality for quadratic forms in subgaussian random ve…

Uniform deviation bounds limit the difference between a model's expected loss and its loss on an empirical sample uniformly for all models in a learning problem. As such, they are a critical component to empirical risk minimization. In this paper, we provide a novel framework to obtain uniform deviation bounds for loss…

We solve robust regression and matrix completion problems with sparse and low-rank models.

Study reduces human labeling in LLM-based classification systems.

We study the problem of high-dimensional sparse mean estimation in the presence of an $ε$-fraction of adversarial outliers. Prior work obtained sample and computationally efficient algorithms for this task for identity-covariance subgaussian distributions. In this work, we develop the first efficient algorithms for rob…

We study an extention of total variation denoising over images to over Cartesian power graphs and its applications to estimating non-parametric network models. The power graph fused lasso (PGFL) segments a matrix by exploiting a known graphical structure, $G$, over the rows and columns. Our main results shows that for …

Analysis of non-asymptotic estimation error and structured statistical recovery based on norm regularized regression, such as Lasso, needs to consider four aspects: the norm, the loss function, the design matrix, and the noise model. This paper presents generalizations of such estimation error analysis on all four aspe…

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…

I introduce and analyse an anytime version of the Optimally Confident UCB (OCUCB) algorithm designed for minimising the cumulative regret in finite-armed stochastic bandits with subgaussian noise. The new algorithm is simple, intuitive (in hindsight) and comes with the strongest finite-time regret guarantees for a hori…

New TS algorithms improve performance in non-stationary multi-armed bandit problems.