Directional differentiability for supremum-type functionals: statistical applications

DOI10.3150/19-BEJ1188zbMATH Open1442.62110arXiv1902.01136OpenAlexW3019987585MaRDI QIDQ2174996FDOQ2174996

Authors: Javier Cárcamo, Luis-Alberto Rodríguez, Antonio Cuevas 

Publication date: 27 April 2020

Published in: Bernoulli (Search for Journal in Brave)

Abstract: We show that various functionals related to the supremum of a real function defined on an arbitrary set or a measure space are Hadamard directionally differentiable. We specifically consider the supremum norm, the supremum, the infimum, and the amplitude of a function. The (usually non-linear) derivatives of these maps adopt simple expressions under suitable assumptions on the underlying space. As an application, we improve and extend to the multidimensional case the results in cite{Raghavachari} regarding the limiting distributions of Kolmogorov-Smirnov type statistics under the alternative hypothesis. Similar results are obtained for analogous statistics associated with copulas. We additionally solve an open problem about the Berk-Jones statistic proposed by cite{Jager-Wellner-2004}. Finally, the asymptotic distribution of maximum mean discrepancies over Donsker classes of functions is derived.

Full work available at URL: https://arxiv.org/abs/1902.01136

Recommendations

• Hadamard differentiability on \(D[0,1]^p\)
• scientific article; zbMATH DE number 1263799
• Second order Hadamard differentiability in statistical applications.
• scientific article; zbMATH DE number 4155679
• A modified functional delta method and its application to the estimation of risk functionals

zbMATH Keywords

Kolmogorov-Smirnov statisticcopulasempirical processesKolmogorov distanceHadamard directional derivativemaximum mean discrepancyDelta methodBerk-Jones statisticKuiper statistic

Mathematics Subject Classification ID

Asymptotic properties of nonparametric inference (62G20) Foundations and philosophical topics in statistics (62A01) Asymptotic distribution theory in statistics (62E20) Characterization and structure theory for multivariate probability distributions; copulas (62H05)

Cites Work

• The earth mover's distance as a metric for image retrieval
• Weak convergence and empirical processes. With applications to statistics
• Asymptotic Statistics
• Asymptotics of empirical copula processes under non-restrictive smoothness assumptions
• Robust Statistics
• On the empirical estimation of integral probability metrics
• Goodness-of-fit tests via phi-divergences
• Mathematical foundations of infinite-dimensional statistical models
• Integral Probability Metrics and Their Generating Classes of Functions
• A kernel two-sample test
• On Weak Convergence of Stochastic Processes with Multidimensional Time Parameter
• Optimal Transport
• Title not available (Why is that?)
• Convergence Criteria for Multiparameter Stochastic Processes and Some Applications
• Asymptotic theory of statistics and probability
• An overview of the goodness-of-fit test problem for copulas
• Title not available (Why is that?)
• Uniform Central Limit Theorems
• Functional analysis, Sobolev spaces and partial differential equations
• Goodness-of-fit test statistics that dominate the Kolmogorov statistics
• Title not available (Why is that?)
• On tests of radial symmetry for bivariate copulas
• Nonparametric tests of density ratio ordering
• On the ``Poisson boundaries of the family of weighted Kolmogorov statistics
• A general limit theorem for recursive algorithms and combinatorial structures
• On nondifferentiable functions and the bootstrap
• On the optimal estimation of probability measures in weak and strong topologies
• Convergence de la répartition empirique vers la répartition théorique
• Tests of stochastic monotonicity with improved power
• On concepts of directional differentiability
• The methods of distances in the theory of probability and statistics.
• Asymptotic analysis of stochastic programs
• Similarity of samples and trimming
• On the contraction method with degenerate limit equation.
• Bracketing metric entropy rates and empirical central limit theorems for function classes of Besov- and Sobolev-type
• A continuous mapping theorem for the smallest argmax functional
• Limiting distributions of Kolmogorov-Smirnov type statistics under the alternative
• Probability Metrics
• Essential supremum norm differentiability
• Supremum norm differentiability
• Equivalence of regression curves
• A contamination model for the stochastic order
• Inference on directionally differentiable functions
• Inference for empirical Wasserstein distances on finite spaces
• Functional data analysis in the Banach space of continuous functions
• New distance measures for classifying X-ray astronomy data into stellar classes
• Integrated empirical processes in \(L^{p}\) with applications to estimate probability metrics
• Weak convergence of the least concave majorant of estimators for a concave distribution function
• A dual approach to inference for partially identified econometric models
• Linear Interpolation With a Nonparametric Accelerated Failure-Time Model
• Title not available (Why is that?)

Cited In (17)

• Two-sample goodness-of-fit tests on the flat torus based on Wasserstein distance and their relevance to structural biology
• Box-constrained monotone approximations to Lipschitz regularizations, with applications to robust testing
• Testing equivalence of multinomial distributions -- a constrained bootstrap approach
• Detecting relevant differences in the covariance operators of functional time series: a sup-norm approach
• A unifying approach to distributional limits for empirical optimal transport
• Limit distribution theory for smooth \(p\)-Wasserstein distances
• Statistical inference for generative adversarial networks and other minimax problems
• Central limit theorems for semi-discrete Wasserstein distances
• Title not available (Why is that?)
• On distributions with fixed marginals maximizing the joint or the prior default probability, estimation, and related results
• On approximate validation of models: a Kolmogorov-Smirnov-based approach
• Hadamard differentiability on \(D[0,1]^p\)
• Empirical optimal transport under estimated costs: distributional limits and statistical applications
• Statistical inference with regularized optimal transport
• Uniform inference for value functions
• Measuring non-exchangeable tail dependence using tail copulas
• A uniform kernel trick for high and infinite-dimensional two-sample problems

Uses Software

• EMD

This page was built for publication: Directional differentiability for supremum-type functionals: statistical applications

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2174996)