Nonparametric estimation of functionals of the derivatives of a signal observed in white Gaussian noise (Q1124244)

Suppose that we observe on a finite interval an integral of a signal S through a channel corrupted by an additive white noise. The paper considers the problem of nonparametric estimation of a functional of S and its derivatives. Sufficient conditions for mean square and strong convergence are given. These conditions are shown to be almost necessary. Consistency conditions are expressed in terms of smoothness of the signal and the functional.