Stein's method in high dimensions with applications (Q1951509)

Using Stein's method, the paper provides a bound for the mathematical expectation difference \(\operatorname{E}h(X)-\operatorname{E}h(Z)\), where \(X\) is a multivariable random vector or for which the components are not necessarily independent, \(Z\) is a Gaussian multi-variable random vector and \(h(\,.\,)\) is a given real-valued function. This result could be useful to approximate random variables by Gaussian random variables.