Asymptotic expansions in the Poisson limit theorem (Q1091034)

It is shown how an adaptation of the Stein-Chen approach can be used with a minimum of difficulty to obtain asymptotic expansions for the distribution of sums of independent nonnegative integer random variables in the neighbourhood of the Poisson distribution together with explicit estimates of the truncation error. It is shown as well that these error estimates perform well in comparison with other estimates that are available. Expansions are also obtained for the expectations of at most polynomially growing functions of the sums, which can be used to derive nonuniform estimates of the rate of convergence. Applications to the Poisson binomial and Poisson negative binomial approximations are considered.