Title :
Exact nonlinear filtering via Gauss transform eigenfunctions
Author :
Paldi, Evtan ; Nehorai, Arve
Author_Institution :
Dept. of Electr. Eng., Yale Univ., New Haven, CT, USA
Abstract :
By imposing mild restrictions on the analytic form of the probability density functions (PDFs) characterizing the model, the family of filtering problems for which exact recursive solutions are known is extended. It is shown that the recursive computation can be done with bounded complexity per iteration on the new family. This family has the potential to model many filtering problems
Keywords :
eigenvalues and eigenfunctions; filtering and prediction theory; probability; transforms; Gauss transform eigenfunctions; PDF; bounded complexity per iteration; exact nonlinear filtering; exact recursive solutions; probability density functions; Biological system modeling; Biology computing; Difference equations; Differential equations; Eigenvalues and eigenfunctions; Filtering; Gaussian processes; Integral equations; Probability density function; State estimation;
Conference_Titel :
Acoustics, Speech, and Signal Processing, 1991. ICASSP-91., 1991 International Conference on
Conference_Location :
Toronto, Ont.
Print_ISBN :
0-7803-0003-3
DOI :
10.1109/ICASSP.1991.150181