Navigation:  Menu Commands > Learning Menu > Options >

 

No Smoothed States For Belief Kalman Filter

 

Print this Topic Previous pageReturn to chapter overviewNext page

The joint Kalman filter for computing the belief coefficients of the PLM and the unobserved state variables will by default use smooth estimates of the state variables in the belief coefficient step of the algorithm. An alternative is to use filter estimates of the second lag of the state variables. This option only affects the estimates of the second lag variables in the PLM, but may be very important for the end result. The default behavior in YADA is to use the smoothed estimates of lag two as these are based on the same information as the lag one estimates of the state variables.

 

Additional Information

A detailed discussion about the joint Kalman filter for the belief coefficients and the state variables can be found in Section 17.5 of the YADA Manual.

 

 


Page url: http://www.texlips.net/yada/help/index.html?no_smoothed_states_for_belief_.htm