Model shrinking of wavelet coefficients and applications
Marina Vannucci, Fabio Corradi · Kent Academic Repository (University of Kent) · 1996
This paper refers to a Bayesian approach to the wavelet shrinkage technique. In contrast with other models recently proposed, the model adopted does not assume the totally independence of denoised wavelet coefficients. Theoretical results regarding the covariance structure of wavelet coefficients are presented and suggest possible specifications of the model. Applications to simulated signals are presented.