DocumentCode
1354873
Title
Least-informative Bayesian prior distributions for finite samples based on information theory
Author
Spall, James C. ; Hill, Stacy D.
Author_Institution
Appl. Phys. Lab., Johns Hopkins Univ., Laurel, MD, USA
Volume
35
Issue
5
fYear
1990
fDate
5/1/1990 12:00:00 AM
Firstpage
580
Lastpage
583
Abstract
A procedure is presented, based on Shannon information theory, for producing least-informative prior distributions for Bayesian estimation and identification. This approach relies on constructing an optimal mixture distribution and applies in small sample sizes (unlike certain approaches based on asymptotic theory). The procedure is illustrated in a small-scale numerical study and is contrasted with an approach based on maximum entropy
Keywords
Bayes methods; estimation theory; identification; information theory; Bayesian estimation; Shannon; finite samples; identification; information theory; least-informative prior distributions; maximum entropy; Bayesian methods; Computer aided software engineering; Control systems; Error correction; H infinity control; Information theory; Optimal control; Process design; Steady-state; Thumb;
fLanguage
English
Journal_Title
Automatic Control, IEEE Transactions on
Publisher
ieee
ISSN
0018-9286
Type
jour
DOI
10.1109/9.53528
Filename
53528
Link To Document