AbstractThe composite lognormal distribution is a popular model for among others insurance data. This thesis is about theory and applications of the composite lognormal distribution. We derive among others the Fisher information matrix and a discrete version of the composite lognormal distribution. We also attempt to give a bivariate and a multivariate version of the composite lognormal distribution. Applications discussed use Canadian net wealth data, cumulative coal production data and insurance data.
|Date of Award
|1 Aug 2022
|Peter Foster (Supervisor) & Saraleesan Nadarajah (Supervisor)
- Bivariate and multivariate distribution
- Insurance data
- Composite distribution
- Lognormal distribution