On the Gaussian approximation for master equations (Q1958574)

Motivated by van Kampen's expansion approach, the authors propose a method to obtain the first and second moments of a stochastic process described by a master equation, relying on an expansion in a large parameter \(\Omega\) of the system, typically the volume. They consider a Gaussian approximation scheme, considering from the very beginning that the fluctuations are Gaussian. By means of a theoretical analysis and of the numerical study of specific examples they compare how their method performs with respect to van Kampen's expansion to first order, looking in particular at the scaling with \(\Omega\) of the error in mean value, second moment and variance.