Heterogeneous Weibull count distribution

被引：1

作者：

Karina, Tania M. ^{[1
]}

Nurrohmah, Siti ^{[1
]}

Fithriani, Ida ^{[1
]}

机构：

[1] Univ Indonesia, Dept Math, Depok 16424, Indonesia

来源：

INTERNATIONAL CONFERENCE ON MATHEMATICS: PURE, APPLIED AND COMPUTATION | 2019年 / 1218卷

关键词：

MODELS;

D O I：

10.1088/1742-6596/1218/1/012020

中图分类号：

O29 [应用数学];

学科分类号：

070104 ;

摘要：

In this paper, we present a generalized model for count data based upon an assumed Weibull interarrival times. Weibull interarrival times could handle overdispersed data with shape parameter 0 < c < 1, and underdispersed data with c > 1, and is reduced as exponential when c = 1. Weibull count model will be obtained by Taylor expansion and convolution method. By mixing, we obtain heterogeneous Weibull count model. Finally, we fit this model and several other models to non-equidispersed data and show that this model fit the data better than Poisson.

引用

页数：10

共 6 条

[1] On the credibility of insurance claim frequency: Generalized count models and parametric estimators [J].