scientific article; zbMATH DE number 3888610

From MaRDI portal
Publication:3221105





The main result in this paper is a normal approximation of a counting process with independent increments. Let (N,\({\mathbb{F}})\) be a counting process with independent increments and denote by A the expectation process and by V the variance process of N. Let \(\Phi\) be the standard normal distribution function. If \(V_ t>0\), then we have \[ (1) \sup_{x}| P(N_ t\leq x)-\Phi ((x-A_ t)/(V_ t)^{1/2})| \leq K V_ t^{-1/8}. \] We can take \(K\) to be \(= (32)/((\pi)(2\pi)^{3/8}(18)^{1/4})\) in the general case and if \(A\) is continuous, then we can replace 18 by 72 in the constant \(K\). These values of \(K\) are better than the ones mentioned in the paper. We also study the probability structure of a counting process with independent increments. We give an alternative way to obtain the distributions of the jump times and inter-arrival times of the process.











This page was built for publication:

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3221105)