Bayesian estimation of non-stationary Markov models combining micro and macro data

In this poster a Bayesian estimation framework for a non-stationary Markov model is developed for situations where sample data with observed transition between classes (micro data) and aggregate population shares (macro data) are available. Posterior distributions on transition probabilities are derived based on a micro based prior and a macro based Likelihood function thereby consistently combining previously separated approaches. Monte Carlo simulations for ordered and unordered Markov states show how observed micro transitions improve precision of posterior knowledge as the sample size increases.


Issue Date:
Jul 24 2011
Publication Type:
Conference Paper/ Presentation
PURL Identifier:
http://purl.umn.edu/103645
Total Pages:
2
Series Statement:
Poster
ID12782




 Record created 2017-04-01, last modified 2017-08-26

Fulltext:
Download fulltext
PDF

Rate this document:

Rate this document:
1
2
3
 
(Not yet reviewed)