## Bernoulli

- Bernoulli
- Volume 2, Number 4 (1996), 341-363.

### Exponential convergence of Langevin distributions and their discrete approximations

Gareth O. Roberts and Richard L. Tweedie

#### Abstract

In this paper we consider a continuous-time method of approximating a given distribution using the Langevin diffusion . We find conditions under which this diffusion converges exponentially quickly to or does not: in one dimension, these are essentially that for distributions with exponential tails of the form , , exponential convergence occurs if and only if . We then consider conditions under which the discrete approximations to the diffusion converge. We first show that even when the diffusion itself converges, naive discretizations need not do so. We then consider a 'Metropolis-adjusted' version of the algorithm, and find conditions under which this also converges at an exponential rate: perhaps surprisingly, even the Metropolized version need not converge exponentially fast even if the diffusion does. We briefly discuss a truncated form of the algorithm which, in practice, should avoid the difficulties of the other forms.

#### Article information

**Source**

Bernoulli Volume 2, Number 4 (1996), 341-363.

**Dates**

First available in Project Euclid: 4 May 2007

**Permanent link to this document**

https://projecteuclid.org/euclid.bj/1178291835

**Mathematical Reviews number (MathSciNet)**

MR1440273

**Zentralblatt MATH identifier**

0870.60027

**Keywords**

diffusions discrete approximations geometric ergodicity Hastings algorithms irreducible Markov processes Langevin models Markov chain Monte Carlo Metropolis algorithms posterior distributions

#### Citation

Roberts, Gareth O.; Tweedie, Richard L. Exponential convergence of Langevin distributions and their discrete approximations. Bernoulli 2 (1996), no. 4, 341--363. https://projecteuclid.org/euclid.bj/1178291835.