The project aims at the improvement of the filtering techniques used to discriminate between trends and cycles in economic time
series and also at the development of time-frequency (local Fourier bases) and time-scale methods (wavelet bases) for non-stationary
data.