Title: Weak consistency of Markov chain Monte Carlo methods

Abstract: Markov chain Monte Calro methods (MCMC) are commonly used in Bayesian statistics. In the last twenty years, many results have been established for the calculation of the exact convergence rate of MCMC methods. We introduce another rate of convergence for MCMC methods by approximation techniques. This rate can be obtained by the convergence of the Markov chain to a diffusion process. We apply it to a simple mixture model and obtain its convergence rate. Numerical simulations are performed to illustrate the effect of the rate.