Abstract
Censored regression models, so-called Tobit models (after Tobin, 1958), occur frequently in econometric and biometric applications. These models play an important role when restrictions for the sampling mechanism are given, for example on the data range.The Bayesian approach is usually very complicated, because closed form solutions for censored and truncated regression models are rarely possible. The non-conjugate situation is treated in two steps. After performing a conjugate analysis, the posterior for the restricted likelihood is derived by using the Gibbs sampling approach.A biologic example to study the growth behaviour of plants is discussed within the context of Tobit models.KeywordsGibbs SamplingTobit ModelFull Conditional DistributionCensor Regression ModelTobit Regression ModelThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have