Local m-estimation with discontinuous criterion for dependent and limited observations

Myung Hwan Seo, Taisuke Otsu

Research output: Contribution to journalArticle

3 Citations (Scopus)

Abstract

We examine the asymptotic properties of local M-estimators under three sets of high-level conditions. These conditions are sufficiently general to cover the minimum volume predictive region, the conditional maximum score estimator for a panel data discrete choice model and many other widely used estimators in statistics and econometrics. Specifically, they allow for discontinuous criterion functions of weakly dependent observations which may be localized by kernel smoothing and contain nuisance parameters with growing dimension. Furthermore, the localization can occur around parameter values rather than around a fixed point and the observations may take limited values which lead to set estimators. Our theory produces three different nonparametric cube root rates for local M-estimators and enables valid inference building on novel maximal inequalities for weakly dependent observations. The standard cube root asymptotics is included as a special case. The results are illustrated by various examples such as the Hough transform estimator with diminishing bandwidth, the maximum score-type set estimator and many others.

Original languageEnglish
Pages (from-to)344-369
Number of pages26
JournalAnnals of Statistics
Volume46
Issue number1
DOIs
Publication statusPublished - 2018 Feb 1
Externally publishedYes

Fingerprint

M-estimation
Estimator
Dependent
Cube root
Dependent Observations
M-estimator
Discrete Choice Models
Maximal Inequality
Kernel Smoothing
Hough Transform
Nuisance Parameter
Diminishing
Panel Data
Econometrics
Asymptotic Properties
Fixed point
Bandwidth
Observation
Cover
Valid

Keywords

  • Cube root asymptotics
  • Maximal inequality
  • Mixing process
  • Parameter-dependent localization
  • Partial identification

ASJC Scopus subject areas

  • Statistics and Probability
  • Statistics, Probability and Uncertainty

Cite this

Local m-estimation with discontinuous criterion for dependent and limited observations. / Seo, Myung Hwan; Otsu, Taisuke.

In: Annals of Statistics, Vol. 46, No. 1, 01.02.2018, p. 344-369.

Research output: Contribution to journalArticle

@article{73f6a7b74a7f495285197fc1f254ef45,
title = "Local m-estimation with discontinuous criterion for dependent and limited observations",
abstract = "We examine the asymptotic properties of local M-estimators under three sets of high-level conditions. These conditions are sufficiently general to cover the minimum volume predictive region, the conditional maximum score estimator for a panel data discrete choice model and many other widely used estimators in statistics and econometrics. Specifically, they allow for discontinuous criterion functions of weakly dependent observations which may be localized by kernel smoothing and contain nuisance parameters with growing dimension. Furthermore, the localization can occur around parameter values rather than around a fixed point and the observations may take limited values which lead to set estimators. Our theory produces three different nonparametric cube root rates for local M-estimators and enables valid inference building on novel maximal inequalities for weakly dependent observations. The standard cube root asymptotics is included as a special case. The results are illustrated by various examples such as the Hough transform estimator with diminishing bandwidth, the maximum score-type set estimator and many others.",
keywords = "Cube root asymptotics, Maximal inequality, Mixing process, Parameter-dependent localization, Partial identification",
author = "Seo, {Myung Hwan} and Taisuke Otsu",
year = "2018",
month = "2",
day = "1",
doi = "10.1214/17-AOS1552",
language = "English",
volume = "46",
pages = "344--369",
journal = "Annals of Statistics",
issn = "0090-5364",
publisher = "Institute of Mathematical Statistics",
number = "1",

}

TY - JOUR

T1 - Local m-estimation with discontinuous criterion for dependent and limited observations

AU - Seo, Myung Hwan

AU - Otsu, Taisuke

PY - 2018/2/1

Y1 - 2018/2/1

N2 - We examine the asymptotic properties of local M-estimators under three sets of high-level conditions. These conditions are sufficiently general to cover the minimum volume predictive region, the conditional maximum score estimator for a panel data discrete choice model and many other widely used estimators in statistics and econometrics. Specifically, they allow for discontinuous criterion functions of weakly dependent observations which may be localized by kernel smoothing and contain nuisance parameters with growing dimension. Furthermore, the localization can occur around parameter values rather than around a fixed point and the observations may take limited values which lead to set estimators. Our theory produces three different nonparametric cube root rates for local M-estimators and enables valid inference building on novel maximal inequalities for weakly dependent observations. The standard cube root asymptotics is included as a special case. The results are illustrated by various examples such as the Hough transform estimator with diminishing bandwidth, the maximum score-type set estimator and many others.

AB - We examine the asymptotic properties of local M-estimators under three sets of high-level conditions. These conditions are sufficiently general to cover the minimum volume predictive region, the conditional maximum score estimator for a panel data discrete choice model and many other widely used estimators in statistics and econometrics. Specifically, they allow for discontinuous criterion functions of weakly dependent observations which may be localized by kernel smoothing and contain nuisance parameters with growing dimension. Furthermore, the localization can occur around parameter values rather than around a fixed point and the observations may take limited values which lead to set estimators. Our theory produces three different nonparametric cube root rates for local M-estimators and enables valid inference building on novel maximal inequalities for weakly dependent observations. The standard cube root asymptotics is included as a special case. The results are illustrated by various examples such as the Hough transform estimator with diminishing bandwidth, the maximum score-type set estimator and many others.

KW - Cube root asymptotics

KW - Maximal inequality

KW - Mixing process

KW - Parameter-dependent localization

KW - Partial identification

UR - http://www.scopus.com/inward/record.url?scp=85042492621&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85042492621&partnerID=8YFLogxK

U2 - 10.1214/17-AOS1552

DO - 10.1214/17-AOS1552

M3 - Article

VL - 46

SP - 344

EP - 369

JO - Annals of Statistics

JF - Annals of Statistics

SN - 0090-5364

IS - 1

ER -