|
Activity Number:
|
453
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 6, 2008 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Risk Analysis
|
| Abstract - #300621 |
|
Title:
|
Sensitivity to Distance, Baseline Distributions, and Forecast Evaluation in Decision and Risk Analysis
|
|
Author(s):
|
Victor R. Jose*+ and Robert Nau and Robert Winkler
|
|
Companies:
|
Duke University and Duke University and Duke University
|
|
Address:
|
Box 90120, Durham, NC, 27708-0120,
|
|
Keywords:
|
Probability elicitation and verification ; Scoring rules ; Probabilistic risk analysis ; Expert opinion
|
|
Abstract:
|
Assessing the quality of probabilistic forecasts through scoring rules as been an important area of study in the field of decision and risk analysis. In this paper, we construct rich families of scoring rules that are strictly proper (thereby encouraging truthful reporting), sensitive to distance (thereby taking into account ordering of events), and incorporate a baseline distribution relative to which the value of a forecast is measured. We utilize the intuition from the creation of the ranked probability score and extend it from being based on a quadratic rule to any strictly proper scoring rule. In particular through the use of the power and pseudospherical scoring rules, we can create a rich family of scoring rules with the said properties. Areas of possible applications include engineering, meteorology, economics and psychology.
|
- 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 2008 program |
