Rates of Convergence of the Diaconis - Holmes - Neal Markov Chain Sampler with a V-Shaped Stationary Probability

#### M. Hildebrand

2004, v.10, Issue 4, 687-704

ABSTRACT

The Diaconis - Holmes - Neal Markov chain sampler is a modification of the Metropolis algorithm based on the nearest neighbor random walk. For certain V-shaped stationary probabilities $\pi$ and a choice of a parameter $\theta$, this paper proves that the Diaconis - Holmes - Neal Markov chain approaches its stationary distribution faster than the Markov chain of the corresponding Metropolis algorithm.

Keywords: Metropolis algorithm,Diaconis - Holmes - Neal sampler