Similarity of samples and trimming

DOI10.3150/11-BEJ351zbMATH Open1239.62005arXiv1205.1950OpenAlexW3101838018MaRDI QIDQ418241FDOQ418241

Authors: Pedro C. Álvarez-Esteban, Carlos Matrán, Eustasio Del Barrio, Juan Antonio Cuesta-Albertos 

Publication date: 28 May 2012

Published in: Bernoulli (Search for Journal in Brave)

Abstract: We say that two probabilities are similar at level $alpha$ if they are contaminated versions (up to an $alpha$ fraction) of the same common probability. We show how this model is related to minimal distances between sets of trimmed probabilities. Empirical versions turn out to present an overfitting effect in the sense that trimming beyond the similarity level results in trimmed samples that are closer than expected to each other. We show how this can be combined with a bootstrap approach to assess similarity from two data samples.

Full work available at URL: https://arxiv.org/abs/1205.1950

Recommendations

• Sampling pairwise comparison
• Similarity and denoising
• Similarity Recovery From Threshold-Based Sampling Under General Conditions
• Trimmed Comparison of Distributions
• Matching random samples in many dimensions
• Matching samples of multiple views
• Similar component analysis

zbMATH Keywords

bootstrapasymptoticsconsistencyrobustnessWasserstein distancemass transportation problemover-fittingsimilarity of distributionstrimmed probability

Mathematics Subject Classification ID

Nonparametric hypothesis testing (62G10) Asymptotic properties of nonparametric inference (62G20) Foundations and philosophical topics in statistics (62A01) Nonparametric statistical resampling methods (62G09)

Cites Work

• Robust Estimation of a Location Parameter
• A general trimming approach to robust cluster analysis
• Trimmed \(k\)-means: An attempt to robustify quantizers
• Title not available (Why is that?)
• Some asymptotic theory for the bootstrap
• Title not available (Why is that?)
• A Robust Version of the Probability Ratio Test
• Concentration inequalities and model selection. Ecole d'Eté de Probabilités de Saint-Flour XXXIII -- 2003.
• Best approximations to random variables based on trimming procedures
• Uniqueness and approximate computation of optimal incomplete transportation plans
• Minimax tests and the Neyman-Pearson lemma for capacities
• Asymptotics for \(L_2\) functionals of the empirical quantile process, with applications to tests of fit based on weighted Wasserstein distances
• Trimmed Comparison of Distributions
• Free boundaries in optimal transport and Monge-Ampère obstacle problems
• The optimal partial transport problem
• Consistency of the \(\alpha \)-trimming of a probability. Applications to central regions
• On the Huber-Strassen theorem
• Least favorable pairs for special capacities
• Limiting distributions of Kolmogorov-Smirnov type statistics under the alternative
• Assessing when a sample is mostly normal
• \(k\)-sample test based on the common area of kernel density estimators

Cited In (14)

• Box-constrained monotone approximations to Lipschitz regularizations, with applications to robust testing
• Detecting relevant changes in the mean of nonstationary processes -- a mass excess approach
• Searching for a common pooling pattern among several samples
• Trimmed Comparison of Distributions
• Models for the assessment of treatment improvement: the ideal and the feasible
• Directional differentiability for supremum-type functionals: statistical applications
• A contamination model for the stochastic order
• On approximate validation of models: a Kolmogorov-Smirnov-based approach
• Interpoint distance based two sample tests in high dimension
• On the topological data analysis extensions and comparisons
• Rates of convergence for partial mass problems
• The empirical cost of optimal incomplete transportation
• Distributionally robust stochastic programs with side information based on trimmings
• Wide consensus aggregation in the Wasserstein space. Application to location-scatter families

This page was built for publication: Similarity of samples and trimming

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q418241)