Open AccessPendugaan Angka Kematian Bayi dengan Menggunakan Model Poisson Bayes Berhirarki Dua-Level
Author(s) -
Nusar Hajarisman,
Alfin Khairi,
Kusman Sadik
Publication year - 2013
Publication title -
mimbar
Language(s) - English
Resource type - Journals
eISSN - 2303-2499
pISSN - 0215-8175
DOI - 10.29313/mimbar.v29i1.366
Subject(s) - small area estimation , poisson distribution , estimator , statistics , estimation , bayes estimator , java , bayes' theorem , bayesian probability , statistical model , computer science , maximum likelihood , econometrics , geography , mathematics , engineering , systems engineering , programming language
Official institutions of national data providers such as the BPS-Statistics Indonesia is required to produce and present the statistical information, as necessary as a form of contributory BPS region in support of regional development policy and planning. There are survey conducted by BPS capability estimation techniques are still limited, due to the resulting estimators have not been able to directly assumed for small areas. In this article we propose the hierarchical Bayesian models, especially for count data which are Poisson distributed, in small area estimation problem. The model was developed by combining concept of generalized linear model and Fay-Herriot model. The results of the development of this model is implemented to estimate the infant mortality rate in Bojonegoro district, East Java Province.