Bayesian Estimation of the Ratio of Two Means using Dirichlet Process Prior

• 발행기관: 한국자료분석학회
• 저자명: Kim, Hee-Jae Won, Ho-Yon
• 간행물 정보: 『Journal of The Korean Data Analysis Society (JKDAS)』Vol.3 No.4, 387~395쪽, 전체 9쪽
• 주제분류: 자연과학 > 통계학
• 파일형태: PDF
• 발행일자: 2001.12.30

영문 초록

In the past, there have been computational difficulties with the model which adapted the Dirichlet process prior. However, using the MCMC, computational difficulties can be solved. In this paper, the populations of interest is the ratio of two exponential populations and two Poisson populations. The Dirichlet process prior is used to provide a nonparametric Bayesian estimate of a ratio of two means. The proposed procedure is applied to some real and artificial examples, and compare it with classical estimator.

목차

1. Introduction

2. Bayesian Estimation of the Ratio of Two Means using Dirichlet Process Prior

3. Numerical Examples

References

키워드

Gibbs Sampler Mixture of Dirichlet Processes Nonparametric Bayes Ratio of Two Means