|
Activity Number:
|
339
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Tuesday, July 31, 2007 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Risk Analysis
|
| Abstract - #308164 |
|
Title:
|
A General Rule To Average a Set of Frequency Statistics
|
|
Author(s):
|
Joseph Kahan*+
|
|
Companies:
|
FM Global
|
|
Address:
|
1151 Boston Providence Turnpike, Norwood, MA, 02062,
|
|
Keywords:
|
frequency ; regression ; heteroskedasticity
|
|
Abstract:
|
For the purpose of the presentation, a frequency statistic is defined by the number of losses divided by the total number of locations from which a loss could occur. Frequency statistics are estimated for a number of groups. In this presentation is proposed the best way of combining these groups to produce an average frequency statistic, where best is defined as being unbiased with a minimum variance. It is posited that a general regression model can form the structure through which three estimators are identified to be possible candidates for obtaining the average frequency of loss over all the loss groups. The best choice depends critically on the heteroskedasticity of the errors. The variances of these estimators depend critically on the correct specification of the model.
|
- The address information is for the authors that have a + after their name.
- Authors who are presenting talks have a * after their name.
Back to the full JSM 2007 program |
