| State Space Models: A Brief History and Some Recent Developments (2007) | |||||||||||||||
Abstract | |||||||||||||||
| Introduction State space models, also termed dynamic models, relate observations y t ; t = 1; 2;:::, on a response variable Y to unobserved "states" or "parameters" # t , t=1,2,..., by an observation model for y t given # t . The states are assumed to follow a Markovian transition model. Gaussian linear state space models are de#ned by a linear observation model and a linear Markovian transition equation y t = z 0 t # t + # t ; t =1; 2; ::: #1# # t = F t # t,1 + # t ; t =1; 2;::: #2# with independent i.i.d. sequences # t # N#0;# 2 #, # t #<F10. | |||||||||||||||
Details der Publikation | |||||||||||||||
| |||||||||||||||