On the geometric ergodicity of hybrid samplers

G. Fort; E. Moulines; G. O. Roberts; J. S. Rosenthal

doi:10.1239/jap/1044476831

On the geometric ergodicity of hybrid samplers

G. Fort
, E. Moulines
, G. O. Roberts
, J. S. Rosenthal

LMC-IMAG
Department of Mathematics and Statistics, Lancaster University
University of Toronto

Research output: Contribution to journal › Article › peer-review

39 Citations (Scopus)

Abstract

In this paper, we consider the random-scan symmetric random walk Metropolis algorithm (RSM) on ℝ^d. This algorithm performs a Metropolis step on just one coordinate at a time (as opposed to the full-dimensional symmetric random walk Metropolis algorithm, which proposes a transition on all coordinates at once). We present various sufficient conditions implying V-uniform ergodicity of the RSM when the target density decreases either subexponentially or exponentially in the tails.

Original language	English
Pages (from-to)	123-146
Number of pages	24
Journal	Journal of Applied Probability
Volume	40
Issue number	1
DOIs	https://doi.org/10.1239/jap/1044476831
Publication status	Published - 1 Mar 2003

Keywords

Geometric ergodicity
Hybrid sampler
Markov chain Monte Carlo

Access to Document

10.1239/jap/1044476831

Cite this

@article{4230ccd5c40248a48ddf16b993d8e55e,

title = "On the geometric ergodicity of hybrid samplers",

abstract = "In this paper, we consider the random-scan symmetric random walk Metropolis algorithm (RSM) on ℝd. This algorithm performs a Metropolis step on just one coordinate at a time (as opposed to the full-dimensional symmetric random walk Metropolis algorithm, which proposes a transition on all coordinates at once). We present various sufficient conditions implying V-uniform ergodicity of the RSM when the target density decreases either subexponentially or exponentially in the tails.",

keywords = "Geometric ergodicity, Hybrid sampler, Markov chain Monte Carlo",

author = "G. Fort and E. Moulines and Roberts, \{G. O.\} and Rosenthal, \{J. S.\}",

year = "2003",

month = mar,

day = "1",

doi = "10.1239/jap/1044476831",

language = "English",

volume = "40",

pages = "123--146",

journal = "Journal of Applied Probability",

issn = "0021-9002",

publisher = "Cambridge University Press",

number = "1",

}

TY - JOUR

T1 - On the geometric ergodicity of hybrid samplers

AU - Fort, G.

AU - Moulines, E.

AU - Roberts, G. O.

AU - Rosenthal, J. S.

PY - 2003/3/1

Y1 - 2003/3/1

N2 - In this paper, we consider the random-scan symmetric random walk Metropolis algorithm (RSM) on ℝd. This algorithm performs a Metropolis step on just one coordinate at a time (as opposed to the full-dimensional symmetric random walk Metropolis algorithm, which proposes a transition on all coordinates at once). We present various sufficient conditions implying V-uniform ergodicity of the RSM when the target density decreases either subexponentially or exponentially in the tails.

AB - In this paper, we consider the random-scan symmetric random walk Metropolis algorithm (RSM) on ℝd. This algorithm performs a Metropolis step on just one coordinate at a time (as opposed to the full-dimensional symmetric random walk Metropolis algorithm, which proposes a transition on all coordinates at once). We present various sufficient conditions implying V-uniform ergodicity of the RSM when the target density decreases either subexponentially or exponentially in the tails.

KW - Geometric ergodicity

KW - Hybrid sampler

KW - Markov chain Monte Carlo

U2 - 10.1239/jap/1044476831

DO - 10.1239/jap/1044476831

M3 - Article

AN - SCOPUS:0037573237

SN - 0021-9002

VL - 40

SP - 123

EP - 146

JO - Journal of Applied Probability

JF - Journal of Applied Probability

IS - 1

ER -

On the geometric ergodicity of hybrid samplers

Abstract

Keywords

Access to Document

Fingerprint

Cite this