arXiv research

New algorithms improve convergence rates for non-log-concave sampling and log-partition estimation.

Semi-supervised EM improves convergence rate with labeled samples.

This article studies the achievable guarantees on the error rates of certain learning algorithms, with particular focus on refining logarithmic factors. Many of the results are based on a general technique for obtaining bounds on the error rates of sample-consistent classifiers with monotonic error regions, in the real…

Improved sampling for Bayesian neural networks reduces vanishing acceptance rates and increases predictive accuracy.

BAMS uses Bayesian sampling to discover AV failures more efficiently and accurately.

We establish a tight characterization of the worst-case rates for the excess risk of agnostic learning with sample compression schemes and for uniform convergence for agnostic sample compression schemes. In particular, we find that the optimal rates of convergence for size-$k$ agnostic sample compression schemes are of…

Rating prediction is an important application, and a popular research topic in collaborative filtering. However, both the validity of learning algorithms, and the validity of standard testing procedures rest on the assumption that missing ratings are missing at random (MAR). In this paper we present the results of a us…

New coin sampling method for Bayesian inference without learning rates.

New algorithms for sampling in constrained domains without learning rates.

Study improves covariance estimation for SGD under Markovian data, matching best rates.

New learning rate approach reveals phase transitions in SGD performance.

The paper studies PCA of probability measures with varying sample sizes and finds optimal convergence rates.

This work extends diffusion models to handle heavy-tailed targets, improving score estimation and sampling guarantees.

Sampling without replacement speeds up optimization in minimax problems.

CR-AIS improves AIS efficiency by constant rate annealing.

RER improves sample complexity by updating in reverse order.

The paper tightens the regret rate for linear bandit problems.

Optimally estimates a functional using nuisance function tuning and sample splitting.

Paper analyzes Nyström regularization for time series forecasting with sequential sub-sampling.

Improved heart rate and activity recognition with low-power wrist sensors.

Study bounds variance modulation function for K-spider distributions.

Paper analyzes convergence rates of two time-scale AC and NAC algorithms.

This short note aims at (re)proving that the symmetrically normalized graph Laplacian $L=\Id - D^{-1/2}WD^{-1/2}$ (from a graph defined from a Gaussian weighting kernel on a sampled smooth manifold) converges towards the continuous Manifold Laplacian when the sampling become infinitely dense. The convergence rate with …

Paper establishes convergence rates and concentration bounds for stochastic approximation and reinforcement learning with Markovian noise.

Private two-sample tests under LDP achieve minimax rates for multinomial and continuous data.

In this paper, we investigate the statistical convergence rate of a Bayesian low-rank tensor estimator. Our problem setting is the regression problem where a tensor structure underlying the data is estimated. This problem setting occurs in many practical applications, such as collaborative filtering, multi-task learnin…

Rate change calculations in the literature involve deterministic methods that measure the change in premium for a given policy. The definition of rate change as a statistical parameter is proposed to address the stochastic nature of the premium charged for a policy. It promotes the idea that rate change is a property o…

The rate of convergence of weighted kernel herding (WKH) and sequential Bayesian quadrature (SBQ), two kernel-based sampling algorithms for estimating integrals with respect to some target probability measure, is investigated. Under verifiable conditions on the chosen kernel and target measure, we establish a near-geom…

MSTGD optimizes gradient descent with stratified sampling for faster convergence.

In this paper we address the following question: Can we approximately sample from a Bayesian posterior distribution if we are only allowed to touch a small mini-batch of data-items for every sample we generate?. An algorithm based on the Langevin equation with stochastic gradients (SGLD) was previously proposed to solv…

A non-trivial probability structure is evident in the binary data extracted from the up/down price movements of very high frequency data such as tick-by-tick data for USD/JPY. In this paper, we analyze the Sony bank USD/JPY rates, ignoring the small deviations from the market price. We then show there is a similar non-…

While active learning offers potential cost savings, the actual data efficiency---the reduction in amount of labeled data needed to obtain the same error rate---observed in practice is mixed. This paper poses a basic question: when is active learning actually helpful? We provide an answer for logistic regression with t…

Stochastic Gradient TreeBoost is often found in many winning solutions in public data science challenges. Unfortunately, the best performance requires extensive parameter tuning and can be prone to overfitting. We propose PaloBoost, a Stochastic Gradient TreeBoost model that uses novel regularization techniques to guar…

This paper analyzes momentum Q-learning with finite-sample guarantees.

GADD accelerates uniform-rate discrete diffusion models by 2 orders of magnitude.

We consider the problem of minimizing a sum of $n$ functions over a convex parameter set $\mathcal{C} \subset \mathbb{R}^p$ where $n\gg p\gg 1$. In this regime, algorithms which utilize sub-sampling techniques are known to be effective. In this paper, we use sub-sampling techniques together with low-rank approximation …

Sharp Lipschitz bounds for flow-matching and diffusion models with optimal sampling rates.

This paper focuses on learning rate analysis of distributed kernel ridge regression for strong mixing sequences. Using a recently developed integral operator approach and a classical covariance inequality for Banach-valued strong mixing sequences, we succeed in deriving optimal learning rate for distributed kernel ridg…

We consider the online one-class collaborative filtering (CF) problem that consists of recommending items to users over time in an online fashion based on positive ratings only. This problem arises when users respond only occasionally to a recommendation with a positive rating, and never with a negative one. We study t…

Many data-fitting applications require the solution of an optimization problem involving a sum of large number of functions of high dimensional parameter. Here, we consider the problem of minimizing a sum of $n$ functions over a convex constraint set $\mathcal{X} \subseteq \mathbb{R}^{p}$ where both $n$ and $p$ are lar…

ABae efficiently computes subset means with expensive predicates using stratified sampling.

Paper develops a new method for distribution regression with indefinite kernels.

Paper improves TD(0) convergence rate with LFA, i.i.d. samples, and averaging.

New study shows neural networks need many samples for training.

Introduces gradient decay in Softmax for better generalization.

New algorithm outperforms existing ones by focusing on mastering rate.

Unified framework for isotropic SG noise in posterior sampling.

Study robust estimation under varying corruption probabilities in data.