Abstract
This paper responds to a number of criticisms of Dempster-Shafer theory made by Judea Pearl. He criticises Dempster-Shafer belief for not obeying the laws of Bayesian belief; however, these laws leas to well-known problems in the face of ignorance, and seem unreasonably restrictive. It is argued that it is not reasonable to expect a measure of belief to obey Pearl's sandwich principle. The standard representation of "if-then" rules in Dempster-Shafer theory, criticised by Pearl, is justified and favorably constrasted with a conditional probability representation.
| Original language | English |
|---|---|
| Pages (from-to) | 345-365 |
| Number of pages | 21 |
| Journal | International Journal of Approximate Reasoning |
| Volume | 6 |
| Issue number | 3 |
| DOIs | |
| Publication status | Published - May 1992 |
| Externally published | Yes |
Keywords
- Bayesian probability
- conditioning
- Dempster-Shafer theory
- if then rules
- lower probability
- the sandwich principle
Fingerprint
Dive into the research topics of 'How much do you believe?'. Together they form a unique fingerprint.Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver