| Summary: | Many physical or biological processes involve variability over both space and time. A large datas,et and the modelling of space, time, and spatio-temporal interaction cause traditional space time methods are limited. This paper presents an approach to space time prediction that achieves dimension reduction and uses a statistical model that is temporally dynamic and spatially descriptive, called space time Kalman filter. The model also
allows a non dinamic spatial component.
Key Words : prediction, filter, optimal prediction, Bayesian inference, orthonormal basis.
|
|---|