Upper bounds for the tail of the compound negative binomial distribution
โ Scribed by E.G. Willmot; X. Lin
- Publisher
- Elsevier Science
- Year
- 1998
- Tongue
- English
- Weight
- 105 KB
- Volume
- 22
- Category
- Article
- ISSN
- 0167-6687
No coin nor oath required. For personal study only.
โฆ Synopsis
A general upper bound for the tail of the compound negative binomial distribution is constructed. By establishing a connection with the individual risk mode the upper bound is seen to be a (possibly degenerate) mixture of tails of gamma distribution. The bound is sharp in that it is an equality in the compound Pascal-exponential case. Two important special cases of the bound are derived. The issue of construction of an optimal upper bound is considered.
๐ SIMILAR VOLUMES
A comparison of two upper bounds for the tails of compound distributions, both defined in terms of a new worse than used (NWU) distribution, reveals that one is sharper in the decreasing failure rate (DFR) case. An inductive argument is employed to construct a lower bound in terms of a new better th