Reweighting samples under covariate shift using a Wasserstein distance criterion - Archive ouverte HAL Access content directly
Journal Articles Electronic Journal of Statistics Year : 2022

Reweighting samples under covariate shift using a Wasserstein distance criterion

(1, 2) , (1, 3)
1
2
3

Abstract

Considering two random variables with different laws to which we only have access through finite size iid samples, we address how to reweight the first sample so that its empirical distribution converges towards the true law of the second sample as the size of both samples goes to infinity. We study an optimal reweighting that minimizes the Wasserstein distance between the empirical measures of the two samples, and leads to an expression of the weights in terms of Nearest Neighbors. The consistency and some asymptotic convergence rates in terms of expected Wasserstein distance are derived, and do not need the assumption of absolute continuity of one random variable with respect to the other. These results have some application in Uncertainty Quantification for decoupled estimation and in the bound of the generalization error for the Nearest Neighbor Regression under covariate shift.
Fichier principal
Vignette du fichier
Reweighting_samples_under_covariate_shift_using_a_Wasserstein_distance_criterion.pdf (977.48 Ko) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-02968059 , version 1 (16-10-2020)
hal-02968059 , version 2 (06-06-2022)

Identifiers

Cite

Julien Reygner, Adrien Touboul. Reweighting samples under covariate shift using a Wasserstein distance criterion. Electronic Journal of Statistics , 2022, 16 (1), pp.3278-3314. ⟨10.1214/21-EJS1974⟩. ⟨hal-02968059v2⟩
235 View
106 Download

Altmetric

Share

Gmail Facebook Twitter LinkedIn More