Determination of Subjective Prior Distributions for Bayesian Analysis of Maintainability Problems,
Abstract
The maintainability engineer is vitally concerned with the prediction of maintenance times. When the design of the product is new and previous data are non-existent the prediction of maintenance times becomes difficult. The paper offers a new, structured approach to establish subjective prior probability distributions until the engineer's judgment is in agreement with the derived distributions. Computer programs were designed to handle two specific tasks: (1) Derivation of a subjective prior distribution via answers to a series of questions. The answers were displayed graphically. (2) Some of the summary output of the first program become input combinations to find the parameters of the log normal distribution which is the most generally used distribution in maintainability. (Modified author abstract)
Document Details
- Document Type
- Technical Report
- Publication Date
- Dec 01, 1973
- Accession Number
- AD0770905
Entities
People
- Chi-yuan Lin
- George J. Schick
Organizations
- University of Southern California