The generalized logarithmic series distribution (Q910802): Difference between revisions

The generalized logarithmic series distribution (GLSD) can asymptotically be considered as a discretized version of the inverse Gaussian distribution. It is shown that its zero centered version \(GLSD_ 0\) is strictly logconvex and infinitely divisible. Using a result on discrete convolution equivalent distributions the asymptotic behaviour of the Lévy measure of the \(GLSD_ 0\) is derived. Finally, an application to risk theory is given.