Advances and Applications in Statistics
Volume 49, Issue 3, Pages 231 - 243
(September 2016) http://dx.doi.org/10.17654/AS049030231 |
|
SADDLE POINT APPROXIMATION FOR THE RANDOM SUM POISSON-BERNOULLI MODEL
O. Al Mutairi Alya, Ameenah Alofi, Mai Alsoubhi, Sara Alradadi and Huda Alblwi
|
Abstract: Approximations are very important because it is sometimes not possible to precisely represent exact representation, while in some cases the exact answer is already obtained but is very difficult to apply, as well the approximations sometimes simplify the analytical treatments. Compared with other asymptotic approximations, saddle point approximations have the advantage of always generating probabilities, being very accurate in the tails of the distribution, and being accurate with small samples, sometimes even with only one observation. In this paper, saddle point approximation methods have been proven to be useful for a range of problems, such as the random sum statistics (Poisson-Bernoulli) model which is very complex model. |
Keywords and phrases: approximations, saddle point approximations, random sum distribution, Poisson-Bernoulli. |
|
Number of Downloads: 400 | Number of Views: 1557 |
|