Bayes DLM 递推方程中未知参数分布假定的理论证明

Theoretical Proof of the Distribution Assumption of Unknown Parameters in Bayes DLM Recurrence Equations

贝叶斯动态线性模型（简称为ＢａｙｅｓＤＬＭ）是一种特殊的预测方法，它以一个动态方程组和先验信息为基础，通过一组递推修正方程预测指标的变化情况。本文给出递推修正方程的一个假定前提的理论证明。

Bayes dynamic linear models(abbreviated as Bayes DLM) is a special of forecasting method, which can forecast the status of index, on the basis of dynamic equations and a priori information, through a set of recurrent modification equations. This article gives theoretical proof of one assumption prerequisite of recurrent modification equations, showing the rationality and strictness of Bayes DLM.