Simultaneous wavelet and spline smoothing of noisy data
Phillip L. Ainsleigh, Charles K. Chui · IEEE International Conference on Acoustics Speech and Signal Processing · 1993
Two data smoothing algorithms using wavelet transforms are proposed. These algorithms take advantage of the time and frequency localization capabilities of wavelets. The first, a finite impulse response (FIR) filtering algorithm, allows real-time smoothing of selected time epochs within each frequency band. The second, a wavelet-based generalized cross validation (GCV) algorithm, provides potentially more optimal smoothing than standard GCV algorithms by allowing for individual smoothing parameters for each frequency band, as opposed to a single smoothing parameter for the overall signal.>