扩展线性支出系统的贝叶斯估计——Gibbs抽样算法的一个应用

Bayesian Estimation of Expanded Linear Expenditure System by Gibbs Sampler

研究了线性支出系统的贝叶斯估计,提出了似然函数的三种等价形式,在条件共轭先验的假设下,推导了该系统贝叶斯估计的Gibbs抽样器,并用我国1999年城镇居民家庭支出数据进行了计算.

The Bayesian estimation of expanded linear expenditure system （ELES） is studied. Three equivalent forms of the likelihood function for ELES are proposed. Under conditionally conjugate priors, a Gibbs sampler to estimate the Bayesian estimation of ELES is derived, and the approach is illustrated with the data of China＇s living expenditure of urban households in 1999.