No abstract

No abstract

No abstract

No abstract

However, for a Gaussian with the same mean and variance as the NB, as n + m, Pdn) = exp [ -( n -( n , h > ) z / ( 2 ( n , h > 2 ) l / [ J ( 2 a ) ( n , h > )ccexp(-n2) Thus, the right tails of the two distributions differ in form, as is readily seen in Fig. 2.As is also evident in Fig. 2, as n + 0 the left tail of the NNB distribution decays more rapidly than the left tail of the equivalent Gaussian. Indeed for n = 0, PSN(0) for the NNB is p S N ( o ) = exp [-g(N)/(l + )l/(l + (nlh))M % exp (-3b)/(nth)'l Y.-H. LEE