Random sequences optimal estimation by using regression and wavelets
O.S. Amosov, Liudmila N. Amosova · 2009
This paper is concerned with random sequences optimal estimation by using regression with unknown type of the regression function and wavelets. The regression and wavelet based estimation algorithms are offered. It is shown that the Bayesian and the alternative regression and wavelet based algorithms provide estimates with the close accuracy. Two examples for linear and nonlinear filtering and prediction problems are considered.