Vector quantile regression: an optimal transport approach

DOI10.1214/15-AOS1401zbMATH Open1381.62239arXiv1406.4643OpenAlexW4210960245MaRDI QIDQ292882FDOQ292882

Authors: Guillaume Carlier, Victor Chernozhukov, Alfred Galichon 

Publication date: 9 June 2016

Published in: The Annals of Statistics (Search for Journal in Brave)

Abstract: We propose a notion of conditional vector quantile function and a vector quantile regression. A emph{conditional vector quantile function} (CVQF) of a random vector $Y$, taking values in $mathbb{R}^d$ given covariates $Z=z$, taking values in $mathbbR{\mathrm{\%}}^k$, is a map $ulongmapsto{Q}_{YmidZ}(u,z)$, which is monotone, in the sense of being a gradient of a convex function, and such that given that vector $U$ follows a reference non-atomic distribution $F_U$, for instance uniform distribution on a unit cube in $mathbb{R}^d$, the random vector $Q_{YmidZ}(U,z)$ has the distribution of $Y$ conditional on $Z=z$. Moreover, we have a strong representation, $Y=Q_{YmidZ}(U,Z)$ almost surely, for some version of $U$. The emph{vector quantile regression} (VQR) is a linear model for CVQF of $Y$ given $Z$. Under correct specification, the notion produces strong representation, $$, for $f(Z)$ denoting a known set of transformations of $Z$, where $$ is a monotone map, the gradient of a convex function, and the quantile regression coefficients $$ have the interpretations analogous to that of the standard scalar quantile regression. As $f(Z)$ becomes a richer class of transformations of $Z$, the model becomes nonparametric, as in series modelling. A key property of VQR is the embedding of the classical Monge-Kantorovich's optimal transportation problem at its core as a special case. In the classical case, where $Y$ is scalar, VQR reduces to a version of the classical QR, and CVQF reduces to the scalar conditional quantile function. An application to multiple Engel curve estimation is considered.

Full work available at URL: https://arxiv.org/abs/1406.4643

Recommendations

• Vector quantile regression beyond the specified case
• Vector-on-function quantile regression for stationary ergodic processes
• On multivariate quantile regression analysis
• Multiple-output quantile regression through optimal quantization
• Conditional quantile estimation through optimal quantization

zbMATH Keywords

Monge-Kantorovich-Breniervector conditional quantile functionvector quantile regression

Mathematics Subject Classification ID

Nonparametric estimation (62G05) Linear inference, regression (62J99) Characterization and structure theory for multivariate probability distributions; copulas (62H05)

Cites Work

• Regression Quantiles
• Estimating the Technology of Cognitive and Noncognitive Skill Formation
• Quantile regression.
• Quantile and probability curves without crossing
• \(M\)-estimation, convexity and quantiles
• Empirical probability plots and statistical inference for nonlinear models in the two-sample case
• On a Geometric Notion of Quantiles for Multivariate Data
• Robust Tests for Heteroscedasticity Based on Regression Quantiles
• Polar factorization and monotone rearrangement of vector‐valued functions
• Title not available (Why is that?)
• Optimal Transport
• Local Linear Quantile Regression
• Multivariate quantiles and multiple-output regression quantiles: from \(L_{1}\) optimization to halfspace depth
• Comonotonic measures of multivariate risks
• Quantile tomography: using quantiles with multivariate data
• Invariance principles for sums of Banach space valued random elements and empirical processes
• Title not available (Why is that?)
• Title not available (Why is that?)
• Nonparametric Estimation of Nonadditive Random Functions
• Existence and uniqueness of monotone measure-preserving maps
• Nonparametric multivariate descriptive measures based on spatial quantiles
• An Approach to Multivariate Covariate-Dependent Quantile Contours With Application to Bivariate Conditional Growth Charts
• On multivariate quantiles under partial orders

Cited In (32)

• Directional bivariate quantiles: a robust approach based on the cumulative distribution function
• Optimal transport: discretization and algorithms
• Robust clustering tools based on optimal transportation
• Center-Outward R-Estimation for Semiparametric VARMA Models
• Conditional sampling with monotone GANs: from generative models to likelihood-free inference
• Optimal experimental design: formulations and computations
• Distribution and quantile functions, ranks and signs in dimension \(d\): a measure transportation approach
• Title not available (Why is that?)
• On the finite-sample performance of measure-transportation-based multivariate rank tests
• Mathematical imaging and surface processing. Abstracts from the workshop held August 21--27, 2022
• Multivariate quantile impulse response functions
• Estimating impulse-response functions for macroeconomic models using directional quantiles
• On the convergence rate of potentials of Brenier maps
• Identification of unobserved distribution factors and preferences in the collective household model
• Title not available (Why is that?)
• Distribution-free tests of multivariate independence based on center-outward quadrant, Spearman, Kendall, and van der Waerden statistics
• Reduced form vector directional quantiles
• Measuring linear correlation between random vectors
• Limit theorems for entropic optimal transport maps and Sinkhorn divergence
• A Bayesian Approach to Multiple-Output Quantile Regression
• Nonlinear inverse optimal transport: identifiability of the transport cost from its marginals and optimal values
• Geometry-Sensitive Ensemble Mean Based on Wasserstein Barycenters: Proof-of-Concept on Cloud Simulations
• Statistical inference with regularized optimal transport
• Flexible quantile contour estimation for multivariate functional data: beyond convexity
• Integrated quantile functions: properties and applications
• Recovering Latent Variables by Matching
• Noncrossing structured additive multiple-output Bayesian quantile regression models
• Geometry of vectorial martingale optimal transportations and duality
• Vector quantile regression beyond the specified case
• Independent Nonlinear Component Analysis
• Multivariate goodness-of-fit tests based on Wasserstein distance
• Multivariate ranks and quantiles using optimal transport: consistency, rates and nonparametric testing

This page was built for publication: Vector quantile regression: an optimal transport approach

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q292882)