The paper advances U-statistics in dependent settings, improving spectral estimation and goodness-of-fit tests.
problem Non-asymptotic analysis of U-statistics in dependent Markov chain settings.
method Proved new concentration and exponential inequalities for U-statistics, applied to spectral estimation, online algorithms, and goodness-of-fit tests.
result Established new results for spectral estimation, online algorithms, and goodness-of-fit tests in Markov chain settings.
Active inference framework improves U-statistic estimation efficiency.
problem Costly acquisition of labels for U-statistics. method Active inference framework with optimal sampling rule.
result Substantial gains in estimation efficiency over baseline methods.
This paper simplifies computing higher-order U-statistics efficiently.
problem The inefficiency of computing higher-order U-statistics in practice. method Decomposition, connection to Einstein summation, and treewidth-based complexity estimate.
result A new, more efficient algorithm to compute U-statistics. Study on U-statistics with heavy-tailed samples, providing tail bounds and LDP.
problem Deviation of U-statistics with heavy-tailed samples.
method Exponential tail bounds and Large Deviation Principle (LDP) for U-statistics.
result Obtained an exponential upper bound for U-statistics tail decay, showing two regions of decay.
Paper develops efficient incomplete U-statistics for degenerate cases.
problem High computational cost and non-standard asymptotic behavior in degenerate U-statistics.
method Characterizes dependence structure using hypergraph theory and combinatorial designs, bypassing traditional Hoeffding decomposition.
result Derives a Berry-Esseen bound for incomplete U-statistics of deterministic designs, enabling Gaussian limiting distributions in degenerate cases.
Study efficient estimation of hidden subspaces in Gaussian Multi-index models.
problem Estimating hidden subspaces in Gaussian Multi-index models with low-dimensional projections.
method Introduced the generative leap exponent and developed an agnostic sequential estimation procedure using spectral U-statistics.
result Achieved optimal sample complexity of $n=Θ(d^{1 \vee \k/2})$ for efficient estimation.
U-statistics improve gradient estimation in importance-weighted variational inference.
problem High variance in gradient estimation for importance-weighted variational inference.
method Use U-statistics to average base gradient estimators on overlapping batches of size m, achieving lower variance.
result U-statistic variance reduction leads to modest to significant improvements in inference performance.
New estimator for symmetric kernel expectations, robust to missing data.
problem Efficient estimation of symmetric kernel expectations with missing data.
method Median-of-Incomplete-U-Statistics (MIU) estimator.
result Established finite-sample concentration rate for MIU.
Enhances U-statistics for semi-supervised datasets using unlabeled data.
problem Efficiently utilizing unlabeled data in semi-supervised settings.
method Semi-supervised U-statistics enhanced by unlabeled data.
result Proposed method is asymptotically Normal and more efficient than classical U-statistics.
Study optimizes KSD estimation from samples, revealing Hilbert-Schmidt vs trace scales.
problem Optimizing estimation of Kernel Stein Discrepancy from samples.
method Identifying and comparing minimax scales for U-statistic and V-statistic.
result Hilbert-Schmidt norm of Stein covariance operator gives optimal scale.
The paper provides bounds for high-dimensional U-statistics with novel order-explicit inequalities.
problem Bounding the deviation of high-dimensional U-statistics from their Hájek projections.
method Develops novel order-explicit moment inequalities for higher-order Hoeffding components.
result The maximum deviation of a high-dimensional U-statistic from its Hájek projection is of order Op(φbn−1log2(dn)). High-dimensional U-statistics show surprising phase transitions, impacting kernel-based tests.
problem Understanding phase transitions in high-dimensional U-statistics.
method Proved a convergence theorem for U-statistics of degree two in high dimensions.
result High-dimensional U-statistics can have non-Gaussian limits with larger variance and asymmetry.
New concentration inequality for U-statistics of Markov chains.
problem Proving a concentration inequality for U-statistics of order two in uniformly ergodic Markov chains.
method Inductive analysis using martingale techniques, uniform ergodicity, Nummelin splitting, and Bernstein's inequality.
result Recovery of convergence rate for U-statistics of independent random variables and canonical kernels, with improved results for dependent kernels.
We revisit resampling procedures for error estimation in binary classification in terms of U-statistics. In particular, we exploit the fact that the error rate estimator involving all learning-testing splits is a U-statistic. Thus, it has minimal variance among all unbiased estimators and is asymptotically normally dis…
Improved estimation of higher order integrals using shrinkage techniques.
problem Estimating higher order Bochner integrals in non-parametric settings.
method Shrinkage of U-statistic towards a target element, considering kernel degeneracy.
result Consistent shrinkage estimators with fast rates of convergence, even for non-degenerate kernels.
Efficient tests for various statistical problems using incomplete U-statistics.
problem Nonparametric tests for two-sample, independence, and goodness-of-fit problems.
method Proposes MMDAggInc, HSICAggInc, and KSDAggInc tests aggregating over multiple kernel bandwidths.
result Aggregated tests provide a solution to the kernel selection problem and achieve optimal rates.
Jackknife variance estimation validated for generalized U-statistics.
problem Uncertainty quantification for subsampling-based estimators.
method Jackknife variance estimation for generalized U-statistics with row-wise Lr weak law. result Jackknife and delete-d variance estimators are ratio-consistent for generalized U-statistics. Unified framework for understanding GRPO as U-statistic.
problem Theoretical properties of GRPO remain less studied.
method Unified framework through classical U-statistics.
result GRPO is asymptotically equivalent to an oracle policy gradient algorithm.
This paper develops dimension-agnostic inference methods for high-dimensional data.
problem Understanding how classical inference methods behave in high-dimensional settings.
method Using variational representations, sample splitting, and self-normalization to create a refined test statistic.
result The resulting statistic has a Gaussian limiting distribution regardless of how dimensionality scales with sample size.
In this paper, we study the problem of computing U-statistics of degree 2, i.e., quantities that come in the form of averages over pairs of data points, in the local model of differential privacy (LDP). The class of U-statistics covers many statistical estimates of interest, including Gini mean difference, Kendal…
Proposes a new method to analyze the distributional effects of treatments.
problem Analyzing the full distributional impact of treatments beyond just the mean.
method Uses kernel conditional mean embeddings and U-statistic regression to investigate the CoDiTE.
result Demonstrates the effectiveness of the proposed method through experiments.
Efficient and robust algorithms for decentralized estimation in networks are essential to many distributed systems. Whereas distributed estimation of sample mean statistics has been the subject of a good deal of attention, computation of U-statistics, relying on more expensive averaging over pairs of observations, is…
There has been an increasing interest in testing the equality of large Pearson's correlation matrices. However, in many applications it is more important to test the equality of large rank-based correlation matrices since they are more robust to outliers and nonlinearity. Unlike the Pearson's case, testing the equality…
A new kernel test avoids permutations for independence testing.
problem Intractable null distributions of kernel statistics.
method Developed xHSIC and xdCov, avoiding permutations.
result New tests have limiting Gaussian distributions under null.
The article introduces practical estimators for kernel discrepancies.
problem Estimating kernel discrepancies accurately and efficiently.
method Presented various estimators for MMD, HSIC, and KSD, including V-statistics, U-statistics, and incomplete U-statistics. Stressed the importance of kernel bandwidth and introduced adaptive estimators.
result Adaptive estimators combining multiple estimators with various kernels address the problem of kernel selection.
Extends conformal prediction for controlling expected risk of monotone loss functions.
problem Controlling expected risk of monotone loss functions.
method Generalizes split conformal prediction with coverage guarantee, extending to distribution shift, quantile risk, multiple, adversarial, and expectations of U-statistics.
result Tight up to an O(1/n) factor, with worked examples in computer vision and natural language processing. New test assesses probabilistic model calibration without expensive approximations.
problem Assessing calibration of probabilistic models with scores.
method Kernel Calibration Conditional Stein Discrepancy (KCCSD) test using new score-based kernels.
result Control over type-I error with improved scalability and efficiency.
New method for MMD with unequal sample sizes improves test power.
problem Existing MMD methods assume equal sample sizes, discarding valuable data.
method Extended generalized U-statistics to handle unequal sample sizes.
result New asymptotic distributions and power optimization for MMD with unequal sample sizes.
USP test improves on Pearson's chi-squared and G-test for independence.
problem Deficiencies in Pearson's chi-squared and G-test for independence. method USP test based on U-statistic estimator of population dependence measure. result USP test controls size, handles small cell counts, and detects minimal violations of independence.
In a wide range of statistical learning problems such as ranking, clustering or metric learning among others, the risk is accurately estimated by U-statistics of degree d≥1, i.e. functionals of the training data with low variance that take the form of averages over k-tuples. From a computational perspective, …
Random forests remain among the most popular off-the-shelf supervised learning algorithms. Despite their well-documented empirical success, however, until recently, few theoretical results were available to describe their performance and behavior. In this work we push beyond recent work on consistency and asymptotic no…
This paper develops a general framework for analyzing asymptotics of V-statistics. Previous literature on limiting distribution mainly focuses on the cases when n→∞ with fixed kernel size k. Under some regularity conditions, we demonstrate asymptotic normality when k grows with n by utilizing existin…
Paper introduces a new measure of conditional dependence avoiding matrix inversions.
problem Measuring conditional dependence between two phenomena influenced by a confounder.
method Uses U-statistics pruning to avoid matrix inversions and re-interpret independence.
result Proposes a novel measure of conditional dependence that avoids matrix inversions.
Improved MoM estimator enhances classical shadows protocol for quantum measurements.
problem Efficient estimation of expectation values with reduced measurement shots.
method Modified median-of-means estimator with optimal constants and U-statistics.
result Improved performance of modified estimator for Clifford measurements.
The method to derive uniform bounds with Gaussian and Rademacher complexities is extended to the case where the sample average is replaced by a nonlinear statistic. Tight bounds are obtained for U-statistics, smoothened L-statistics and error functionals of l2-regularized algorithms.
New unbiased variance estimator for random forests using Hoeffding decomposition.
problem Uncertainty quantification in random forests with large kernel sizes and small sample sizes.
method Proposes a new Hoeffding decomposition view for variance estimation, establishing unbiased estimators and ratio consistency.
result Establishes the ratio consistency of the proposed variance estimator, justifying confidence interval coverage rates.
Unified method for MMD variance estimation improves accuracy and computational efficiency.
problem Variance estimation for MMD in nonparametric testing.
method Unified finite-sample characterization of MMD variance through U-statistic and Hoeffding decomposition; exact acceleration method for univariate case.
result Unified estimators improve accuracy and computational efficiency for MMD variance.
New method for efficient matrix completion with nonignorable missing data.
problem Nonignorable missing data in matrix completion.
method Nuclear norm regularized U-statistic loss function and accelerated proximal gradient algorithm.
result Near minimax optimal statistical convergence rate for nonignorable missing data.
A scalable ROC-SVM variant reduces training time for imbalanced binary classification.
problem High computational cost of ROC-SVM for imbalanced binary classification.
method Incomplete U-statistics and low-rank kernel approximation.
result Comparable AUC performance with reduced training time.
We study the problem of independence testing given independent and identically distributed pairs taking values in a σ-finite, separable measure space. Defining a natural measure of dependence D(f) as the squared L2-distance between a joint density f and the product of its marginals, we first show that there is…
We quantify uncertainty in Oja's algorithm's leading eigenvector estimation.
problem Estimating the error of Oja's algorithm's leading eigenvector from streaming data.
method Combining U-statistics, high-dimensional central limit theorems, and multiplier bootstrap.
result Established a weighted χ² approximation for the error between the eigenvector and algorithm output.
This work develops formal statistical inference procedures for machine learning ensemble methods. Ensemble methods based on bootstrapping, such as bagging and random forests, have improved the predictive accuracy of individual trees, but fail to provide a framework in which distributional results can be easily determin…
We provide bounds for kernel matrices and new approximations for high-dimensional data.
problem Approximating high-dimensional empirical kernel matrices.
method Decoupling results for U-statistics and non-commutative Khintchine inequality.
result New tighter approximations for inner-product kernel matrices.
Paper proposes a method to estimate confidence bands for survival random forests.
problem No statistically valid and computationally feasible approach for estimating confidence bands for survival random forests.
method Extending recent developments in infinite-order incomplete U-statistics, the paper proposes an unbiased confidence band estimation.
result The proposed method accurately estimates the confidence band and achieves desired coverage rate.
This paper quantifies uncertainty in Data Shapley using statistical inference.
problem Uncertainty in data valuation due to dynamic data distribution.
method Established relationship with U-statistics and quantified uncertainty using statistical inference.
result Confidence intervals for Data Shapley estimations are provided.
New tests for high-dimensional data improve on existing methods.
problem Testing mean vectors in high-dimensional data.
method Generalized multivariate sign transformation, using different norm functions.
result Tests using generalized signs have higher power than existing tests.
A family of maximum mean discrepancy (MMD) kernel two-sample tests is introduced. Members of the test family are called Block-tests or B-tests, since the test statistic is an average over MMDs computed on subsets of the samples. The choice of block size allows control over the tradeoff between test power and computatio…
Unified framework for global and local two-sample conditional distribution testing.
problem Testing equality of two conditional distributions.
method Distance and kernel methods, conditional U-statistics, local bootstrap.
result Developed reliable global and local tests.