The Characterization of Strictly Proper Scoring Rules in Decision Making.
Abstract
A strictly proper scoring rule or admissible probability measurement procedure (APMP)is a mathematical device that allows a decision-maker to 'score' probabilistic forecasts made by experts once the outcome is known. The expected score, as seen by the forecaster, should be maximized when the forecaster states his true beliefs so that he will be encouraged to do so. Applications to subjective probability assessment will be given. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Oct 01, 1981
- Accession Number
- ADA110209
Entities
People
- Eduardo Haim
Organizations
- University of California, Berkeley