Abstract
Biomedical signals and systems typically have the following four “N” properties: Nonstationarity, Nonlinearity, Non-Gaussianity, and Nonlong term. In this chapter, some of these properties are addressed by using adaptive signal processing techniques in which digital filter coefficients are designed by using adaptive algorithms. The adaptive algorithms are based on least squares optimization in which the filter coefficients are designed by using minimization of the error square function between the desired response and the estimate of the filter. Linear prediction and time-series modeling are seen as special cases of adaptive filter design in which case the coefficients could be used as features for machine learning applications.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
