The authors introduce a new estimation procedure, Augmented Kalman Fil
ter with Continuous State and Discrete Observations (AKF(C-D)), for es
timating diffusion models, This method is directly applicable to diffe
rential diffusion models without imposing constraints on the model str
ucture or the nature of the unknown parameters. It provides a systemat
ic way to incorporate prior knowledge about the likely Values of unkno
wn parameters and updates the estimates whets new data become availabl
e. The authors compare AKF(G-D) empirically with live other estimation
procedures, demonstrating AKF(C-D)'s superior prediction performance.
As an extension to the basic AKF(C-D) approach, they also develop a p
arallel-filters procedure for estimating diffusion models when there i
s uncertainty about diffusion model structure or prior distributions o
f the unknown parameters.