摘要：Stein’s method was first invented by Stein to study the error of normal approximations for sums of dependent random variables in 1972, and was successfully extended by Chen to Poisson approximations of dependent trials in 1975. Nowadays it has been one of most powerful tools in probability theory and statistical inferences. In this talk I will report some recent applications of Stein’s method, among which are convergence rates of Markov chains induced by Euler Maruyama’s scheme in random regime switching diffusion processes, and error estimates in three-parameter distributional approximation of sums of locally dependent random variables.