Knowledge typology for imprecise probabilities.
- Gregory D.
- Lauren J.
When characterizing the reliability of a complex system there are often gaps in the data available for specific subsystems or other factors influencing total system reliability. At Los Alamos National Laboratory we employ ethnographic methods to elicit expert knowledge when traditional data is scarce. Typically, we elicit expert knowledge in probabilistic terms. This paper will explore how we might approach elicitation if methods other than probability (i.e., Dempster-Shafer, or fuzzy sets) prove more useful for quantifying certain types of expert knowledge. Specifically, we will consider if experts have different types of knowledge that may be better characterized in ways other than standard probability theory.
- Research Organization:
- Los Alamos National Laboratory
- Sponsoring Organization:
- DOE
- OSTI ID:
- 976105
- Report Number(s):
- LA-UR-02-1257
- Country of Publication:
- United States
- Language:
- English
Similar Records
Simplifying Probability Elicitation and Uncertainty Modeling in Bayesian Networks
The case for lower probabilities as measures of uncertainty