Approximate Bayesian Analysis of Censored Survival Data
研究了位置尺度回归模型中右删失数据的贝叶斯分析，发现精度参数和回归系数的后验分布可分别用卡方和多元t分布近似，并用临床生存数据验证。
The Bayesian analysis of data arising from a location-scale regression model, and which may include a number of right-censored observations, is considered. The posterior distributions of the suitably normalized precision parameter and regression coefficient vector are shown to be well approximated by X2 and multivariate t distributions respectively. The ideas are illustrated using some clinical survival data.