定数截尾下Burr-Ⅻ分布参数贝叶斯估计的混合Gibbs算法

Mixed Gibbs Algorithm for Bayesian Estimation of Burr-Ⅻ Distribution Parameters in the Type Ⅱ Censored Sample Occasions

利用混合Gibbs算法（Gibbs抽样与Metropolis算法混合）给出了定数截尾样本场合Burr—XII分布参数的贝叶斯估计，通过Monte-Carlo模拟，考查贝叶斯估计的均值、均方误差及参数的可信区间，并给出混合Gibbs抽样过程中相应参数的轨迹图、直方图及自相关系数图．结果表明：在定数截尾样本场合，用混合Gibbs算法求Burr—XII分布参数的贝叶斯估计得到了比较满意的结果，算法可行、稳定、有效．

It gives Bayesian estimation of Burr-XII distribution parameters using mixed Gibbs algorithm （Gibbs sampling with Metropolis algorithm mixed） in the type Ⅱ censored samples occasions,examines the estimated mean,mean square error and the confidence interval for the parameter by the Monte-Carlo simulation, gives locus,histograms and autocorrelation coefficient figure of the corresponding parameter in the the mixed Gibbs sampling process.The results show that： Bayesian estimation of Burr-XII distribution parameters using mixed Gibbs algorithm in the type Ⅱ censored samples occasions are quite satisfied, the algorithm is a viable, stable and effective.