A Comparison of the Characteristics of Some Bayesian Forecasting Models
对比了几种非线性贝叶斯预测模型，指出它们都能产生一致的预测方法，并给出了一套性质来帮助实践者选择适合的预测方案，最后展示了如何调整动态广义线性模型使其一致。
Several different types of non-linear Bayesian forecasting models are contrasted and compared. Each is shown capable of producing coherent forecasting methodologies. A set of desirable properties is presented to help guide the practitioner in a choice of an appropriate forecasting scheme for his particular application. Some characterizations and impossibility theorems are then given. The paper ends by illustrating how to adapt the Dynamic Generalized Linear Model to make it coherent.