Functional estimation in high-dimensional and infinite-dimensional models
Abstract
Let be a family of probability measures on a measurable space Given a Banach space a functional and a mapping our goal is to estimate based on i.i.d. observations In particular, if is an identifiable statistical model with parameter set one can consider the mapping for resulting in a problem of estimation of based on i.i.d. observations Given a smooth functional and estimators of we use these estimators, the sample split and the Taylor expansion of of a proper order to construct estimators of For these estimators and for a functional of smoothness we prove upper bounds on the -errors of estimator under certain moment assumptions on the base estimators We study the performance of estimators in several concrete problems, showing their minimax optimality and asymptotic efficiency. In particular, this includes functional estimation in high-dimensional models with many low dimensional components, functional estimation in high-dimensional exponential families and estimation of functionals of covariance operators in infinite-dimensional subgaussian models.
Cite
@article{arxiv.2310.16129,
title = {Functional estimation in high-dimensional and infinite-dimensional models},
author = {Vladimir Koltchinskii and Minghao Li},
journal= {arXiv preprint arXiv:2310.16129},
year = {2023}
}