On a two-step estimation of a multivariate logit model
On a two-step estimation of a multivariate logit model
复制标题
多元 Logit 模型的两步估计
DOI：
10.1016/0304-4076(78)90086-6
复制
发表时间：
1978
影响因子：
6.3
通讯作者：
T. Amemiya
中科院分区：
文献类型：
--
作者：
T. Amemiya
本文研究了Domencich和McFadden（Urban Travel Demand，North-Holland，1975）提出的多元Logit模型的两步估计方法的性质，证明了它是相容的，但渐近效率低于极大似然估计。然而，它的计算可以比最大似然估计简单得多，特别是在涉及多个因变量的模型中。
In this article the author studies the properties of the two-step estimation method proposed by Domencich and McFadden (Urban Travel Demand, North-Holland, 1975) for a multivariate logit model and shows that it is consistent but asymptotically less efficient than the maximum likelihood estimator. Its computation, however, can be considerably simpler than that of the maximum likelihood estimator, especially in models involving several dependent variables.