Efficient Sampling of Stochastic Differential Equations with Positive Semi-Definite Models

Neural Information Processing Systems 

This paper deals with the problem of efficient sampling from a stochastic differential equation, given the drift function and the diffusion matrix.