Mollification based on wavelets (Q402300)

In this paper the authors study the mollification of functions using wavelets so as to obtain approximations of derivatives of noisy functions. The kind of mollifiers proposed are obtained from orthogonal as well as non-orthogonal wavelets, in particular, non-orthogonal systems of \(B\)-splines and non-orthogonal Franklin's wavelet. The wavelets are so chosen that noise is reduced and the Gibbs phenomenon does not occur. Estimates of the error in approximation by the mollification are given for the case when the fractional derivative of a function is calculated.