Skip to content
Back to formatted view

Raw Message

Message-ID: <AANLkTikDyZeNhKfncNdC7hp_CveXSafbxmM89V-pTMER@mail.gmail.com>
Date: 2011-01-20T14:51:18Z
From: Artyom Kharitonov
Subject: Kalman and regression model

Hi everyone. I have a question about the Kalman filter and DLM package.
I have Initial data:
regression : y = bx
Purpose: make adaptive b
Decision:
After MLE define the unknown parameter V, W.
Define state-space model dlmRegr (...) with V,W
Than we implement dlmFilter and dlmSmooth (As i know, on this period
dlmFilter estimates b). This method works well for in-sample area.
How  I can implement this approach to adapt b  for future data, using
results obtaining at previous steps?
Thanks