Dear Dongsong,
we have analyzed time series modeling with missing and noisy data
in part from a graphical models point of view in two papers
Tresp, V., Hofmann, R. (1995).
Missing and Noisy Data in Nonlinear Time-Series Prediction
In Girosi, F., Makhoul, J., Manolakos E., Wilson E. (Hrsg.),
{\em Neural Networks for Signal Processing 5,}
IEEE catalog number: 95TH8094.
http://wwwbrauer.informatik.tu-muenchen.de/~hofmannr/miss_data95_abstr.html
Tresp, V., Hofmann, R. (1998).
Nonlinear Time-Series Prediction with Missing and Noisy Data.
{\em Neural Computation,} Vol. 10.
http://wwwbrauer.informatik.tu-muenchen.de/~hofmannr/nc97_abstr.html
Greetings,
- Volker
_______________________________________________________________________________
Volker Tresp
Siemens AG
ZT IK 4
Otto-Hahn-Ring 6
D-81730 Muenchen
Germany
Phone: +49-(0)89-636 49408 Secretariat: +49-(0)89-636 49725
Fax: +49-(0)89-636 49767 E-mail: [EMAIL PROTECTED]
