Further Theory of Stable Decisions.
Abstract
In a decision problem, a Bayesian has a loss function, a likelihood function, and a prior opinion. A Bayesian makes a decision by minimizing expected loss. Assuming the likelihood function is fixed and agreed upon, this thesis examines the influence of small variations in loss function and prior opinion. There are a lot of ways to define small variations in loss function and prior distribution. Essentially one has to define distance between two distributions and distance between two loss functions.
Document Details
- Document Type
- Technical Report
- Publication Date
- Aug 01, 1978
- Accession Number
- ADA060026
Entities
People
- David T. Chuang
Organizations
- Carnegie Mellon University