Back to search
University of New Hampshire
SOME PROBLEMS IN BAYESIAN INFERENCE IN A NONCLASSICAL SETTING (OPERATIONAL STATISTICS, BETTING RATES, COHERENCE)
Abstract
dc:description.abstract<p>The Bayesian inference strategy is studied within the framework of operational statistics (generalized sample spaces). Conditions necessary and sufficient for consistency and strict consistency of a complete betting rate assignment are derived. A constructive proof is given of the existence of a unique second-order probability which induces a given consistent and countably additive complete betting rate assignment. Continuity of the Bayesian inference strategy is investigated. Asymptotic properties of the Bayesian inference strategy are examined both at the level of betting rates and at the level of second-order probabilities.</p>
Degree
thesis:*- Name thesis:degree_name
- Doctor of Philosophy
- Level thesis:degree_level
- Dissertation
- Year
- 1984
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- MAGNESS, CAROLYN MARGARET
Subjects
dc:subject × 1Identifiers
dc:identifier.*- Repository record dc:identifier
- https://scholars.unh.edu/dissertation/1429
- OAI identifier oai:identifier
- oai:scholars.unh.edu:dissertation-2428