On the wavelet denoising of multifractal functions

Aurélia Fraysse (1)

1 L2S - Laboratoire des signaux et systèmes

This paper is devoted to the link between the rate of convergence of functions by thresholding algorithms and the multifractal formalism. Indeed we show that once given its Besov regularity, almost every function in the prevalence sense, is approximated at the minimax rate. We also prove that when the multifractal formalism is fulfilled the corresponding functions are approximated at the minimax rate of convergence.

Domaines: Statistiques [math.ST]; Théorie [stat.TH]

Aurélia Fraysse. On the wavelet denoising of multifractal functions. 2012. ⟨hal-00731464⟩