DocumentCode :
819932
Title :
Minimax and 
curve fitting in non-Gaussian MAP estimation
curve fitting in non-Gaussian MAP estimationAuthor :
Scott, Peter D.
Author_Institution :
State University of New York, Buffalo, NY, USA
Volume :
20
Issue :
5
fYear :
1975
fDate :
10/1/1975 12:00:00 AM
Firstpage :
690
Lastpage :
691
Abstract :
A class of non-Gaussian estimation problems is equivalent to minimax and L1 curve fitting. The curve fit is shown to be algebraically dual to optimization of a positive semidefinite quadratic form with linear inequalities, which is solved by a fast quadratic program based on Graves´ simplex algorithm. An example compares the performance of this estimator with the (suboptimal) minimum mean-squared error (MMSE) estimations generated by quadratic curve fitting.
Keywords :
Curve fitting; Least-squares estimation; Linear systems, stochastic discrete-time; Minimax estimation; Optimization methods; State estimation; Covariance matrix; Curve fitting; Laplace equations; Minimax techniques; Probability density function; Quadratic programming; Random variables; Statistics; White noise; Zinc;
fLanguage :
English
Journal_Title :
Automatic Control, IEEE Transactions on
Publisher :
ieee
ISSN :
0018-9286
Type :
jour
DOI :
10.1109/TAC.1975.1101080
Filename :
1101080
Link To Document :
