Bayesian semiparametric estimation of discrete duration models: an application of the dirichlet process prior
提出一种离散时间持续时间模型的贝叶斯估计方法，利用狄利克雷过程先验处理未观测异质性，相比非参数最大似然估计具有有限样本推断和灵活基线风险的优势，并应用于加拿大新不伦瑞克省的就业持续时间数据。
Abstract This paper proposes a Bayesian estimator for a discrete time duration model which incorporates a non‐parametric specification of the unobserved heterogeneity distribution, through the use of a Dirichlet process prior. This estimator offers distinct advantages over the Nonparametric Maximum Likelihood estimator of this model. First, it allows for exact finite sample inference. Second, it is easily estimated and mixed with flexible specifications of the baseline hazard. An application of the model to employment duration data from the Canadian province of New Brunswick is provided. Copyright © 2001 John Wiley & Sons, Ltd.