Summary |
The purpose of the monograph is not to settle this issue or even to attempt to make a contribution toward its resolution. Rather, it is to demonstrate that there is a well-deserved place for Bayesian statistics in the analyst’s arsenal. This is accomplished by showing that two important actuarial problems can be successfully solved by the Bayesian approach. The first is the general problem of model-based prediction and normally involves three steps. Begin by collecting data on the past behavior of an insurance system. This may, for example, be the ages at death of the holders of a certain type of life insurance policy or the indemnity amounts paid to the members covered by a group insurance policy. The second step is to build a model describing the random process that yielded these observations.
|
