A random-discretization based Monte Carlo sampling method and its applications (Q1610847)

A simple numerical sampling based method is proposed. It is based on random discretization of probability density function with respect to Lebesgue measure and a multivariate version of the inverse probability integral transformation. It requires only the knowledge of the functional form of the density function. Therefore, it is dimension-free, non-iterative and applicable to many multivariate probability distributions. Four simple examples of two or three dimensional problems, two benchmark examples of maximum likelihood estimation and three examples relating to the problem in treating with extremely low probability measure are presented for illustration.