Global rates of convergence of the MLEs of log-concave and s-concave densities

Charles R Doss, Jon A. Wellner

Research output: Contribution to journalArticle

17 Citations (Scopus)

Abstract

We establish global rates of convergence for the Maximum Likelihood Estimators (MLEs) of log-concave and s-concave densities on ℝ. The main finding is that the rate of convergence of the MLE in the Hellinger metric is no worse than n-2/5 when -1 < s <∞ where s = 0 corresponds to the log-concave case. We also show that the MLE does not exist for the classes of s-concave densities with s <-1.

Original languageEnglish (US)
Pages (from-to)954-981
Number of pages28
JournalAnnals of Statistics
Volume44
Issue number3
DOIs
StatePublished - Jun 1 2016

Fingerprint

Log-concave
Maximum Likelihood Estimator
Rate of Convergence
Metric
Maximum likelihood estimator
Rate of convergence

Keywords

  • Bracketing entropy
  • Consistency
  • Empirical processes
  • Global rate
  • Hellinger metric
  • Log-concave
  • S-concave

Cite this

Global rates of convergence of the MLEs of log-concave and s-concave densities. / Doss, Charles R; Wellner, Jon A.

In: Annals of Statistics, Vol. 44, No. 3, 01.06.2016, p. 954-981.

Research output: Contribution to journalArticle

@article{b29728caded64468b6920b242eafd636,
title = "Global rates of convergence of the MLEs of log-concave and s-concave densities",
abstract = "We establish global rates of convergence for the Maximum Likelihood Estimators (MLEs) of log-concave and s-concave densities on ℝ. The main finding is that the rate of convergence of the MLE in the Hellinger metric is no worse than n-2/5 when -1 < s <∞ where s = 0 corresponds to the log-concave case. We also show that the MLE does not exist for the classes of s-concave densities with s <-1.",
keywords = "Bracketing entropy, Consistency, Empirical processes, Global rate, Hellinger metric, Log-concave, S-concave",
author = "Doss, {Charles R} and Wellner, {Jon A.}",
year = "2016",
month = "6",
day = "1",
doi = "10.1214/15-AOS1394",
language = "English (US)",
volume = "44",
pages = "954--981",
journal = "Annals of Statistics",
issn = "0090-5364",
publisher = "Institute of Mathematical Statistics",
number = "3",

}

TY - JOUR

T1 - Global rates of convergence of the MLEs of log-concave and s-concave densities

AU - Doss, Charles R

AU - Wellner, Jon A.

PY - 2016/6/1

Y1 - 2016/6/1

N2 - We establish global rates of convergence for the Maximum Likelihood Estimators (MLEs) of log-concave and s-concave densities on ℝ. The main finding is that the rate of convergence of the MLE in the Hellinger metric is no worse than n-2/5 when -1 < s <∞ where s = 0 corresponds to the log-concave case. We also show that the MLE does not exist for the classes of s-concave densities with s <-1.

AB - We establish global rates of convergence for the Maximum Likelihood Estimators (MLEs) of log-concave and s-concave densities on ℝ. The main finding is that the rate of convergence of the MLE in the Hellinger metric is no worse than n-2/5 when -1 < s <∞ where s = 0 corresponds to the log-concave case. We also show that the MLE does not exist for the classes of s-concave densities with s <-1.

KW - Bracketing entropy

KW - Consistency

KW - Empirical processes

KW - Global rate

KW - Hellinger metric

KW - Log-concave

KW - S-concave

UR - http://www.scopus.com/inward/record.url?scp=84963541503&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84963541503&partnerID=8YFLogxK

U2 - 10.1214/15-AOS1394

DO - 10.1214/15-AOS1394

M3 - Article

C2 - 28966409

AN - SCOPUS:84963541503

VL - 44

SP - 954

EP - 981

JO - Annals of Statistics

JF - Annals of Statistics

SN - 0090-5364

IS - 3

ER -