In recursive methods the construction of
an estimate at time
is based on an estimate from the previous time
and the observations available in the time
. Exponential smoothing and
Yule-Walker equations are examples of recursive algorithms but
by defining a state-space
model one can build a unifying theory of recursive methods with the Kalman
filter as a general (linear) solution of filtering, smoothing and prediction problems.
All routines for Kalman filtering, which will be explained in the following, are part of the times library.
![]() |
MD*TECH Method and Data Technologies |
http://www.mdtech.de mdtech@mdtech.de |