Nonlinear Methods

Nonlinear Methods

Author(s): Vladimir Katkovnik, Karen Egiazarian, Jaakko Astola

Published: 2006

https://doi.org/10.1117/3.660178.ch10

Abstract

The LPA estimates studied in the previous chapters give universal and flexible tools for signals given with noise. However, there are a number of good reasons to go beyond these linear estimators to nonlinear ones. If we look at the estimation problem in general, we see that the selection of an algorithm depends mainly on three hypotheses: 1. Signal model; 2. Observation model; 3. Criterion.