On the Structural Representation for Polynomial-Projecting Predictor Models Based on the Kalman Filter
基于 Kalman 滤波的多项式投影预测模型的结构表示
Summary Harrison and Stevens (1976) give a general class of dynamic linear models which incorporates a stochastic structure for describing model components and which utilizes the Kalman filter algorithm to update the system. This paper derives conditions for the dynamic linear models to yield polynomial-projecting predictors which do not have limiting constraints imposed upon them. Some examples of structurally simple models which satisfy these conditions are given, and the…