DocumentCode
698699
Title
Fast maximum-likelihood sea clutter parameter learning from the output of the envelope detector
Author
Sari, Faruk ; Sari, Nursen ; Mili, Lamine
Author_Institution
TUBITAK, MRC, Gebze, Turkey
fYear
2005
fDate
4-8 Sept. 2005
Firstpage
1
Lastpage
4
Abstract
We develop a fast learning technique to estimate the background statistics parameters from the output of the envelope detector, the inputs of which are multi-component Gaussian Mixture (GM) distributions. We use Fisher Scoring (FS) algorithm, which is Newton based and has fast convergence properties, to solve the log-likelihood minimization problem. Experimental results are given on real radar clutter data.
Keywords
Gaussian distribution; Gaussian processes; learning (artificial intelligence); maximum likelihood detection; mixture models; radar clutter; radar detection; FS algorithm; GM distributions; background statistics parameters; envelope detector; fast convergence properties; fast learning technique; fast maximum-likelihood sea clutter parameter learning; fisher scoring algorithm; log-likelihood minimization problem; multicomponent Gaussian mixture distributions; radar clutter data; Clutter; Envelope detectors; Maximum likelihood estimation; Optimization; Radar; Sea state;
fLanguage
English
Publisher
ieee
Conference_Titel
Signal Processing Conference, 2005 13th European
Conference_Location
Antalya
Print_ISBN
978-160-4238-21-1
Type
conf
Filename
7078292
Link To Document