Abstract
The article deals with robust bayesian estimates. Different statistical models are proposed which yield estimates of the posterior mean which are able to cope with very large claims, misspecified prior information and misspecified distributional assumptions. The methods developed in the paper are especially useful when dealing with catastrophe risks.
Keywords: Bayesian statistics, Robust statistics, Catastrophe risks.
Volume
Porto Cervo, Italy
Year
2000
Categories
Financial and Statistical Methods
Statistical Models and Methods
Bayesian Methods
Actuarial Applications and Methodologies
Data Management and Information
Data Quality
Financial and Statistical Methods
Extreme Event Modeling
Publications
ASTIN Colloquium
