截断非平衡似无关模型的极大仿真似然估计

样本数据缺失和截断是现代统计调查中经常遇到的两个问题，它们在一定程度上影响模型参数估计的准确性和有效性. 该研究首先提出了一个新的截断非平衡似无关回归模型，这个模型能够同时考虑数据缺失和截断的特征；然后基于Geweke-Hajivassiliou-Keane（GHK）的仿真算子，建立了该模型的极大仿真似然估计方法；蒙特卡罗实验结果表明，在大样本和有限样本下这种估计方法在参数估计的准确性和有效性方面均具有良好表现.

Maximum simulated likelihood estimation of censored seemingly unrelated regressions with missing observations

In many survey data sets, multiple observations are frequently missed and censored at some threshold values, which cause the biased estimation based on such samples. In this paper, we develop a censored and unbalanced seemingly unrelated regression model, allowing for the censored dependent variables and missing observations. By using the Geweke-Hajivassiliou-Keane (GHK) simulator, we propose a procedure of maximum simulated likelihood estimators for this new model. A small scale Monte Carlo simulation demonstrates that the estimators can perform well both in finite and large sample.