Wavelet and Fourier regularization methods are effective for the nonparametric regression problem. We prove that the loss function evaluated for the regularization parameter chosen through GCV or Mallows criteria is asymptotically equivalent in probability to its minimum over the regularization parameter. © 2001 Elsevier Science B.V.
Convergence in probability of the Mallows and GCV wavelet and Fourier regularization methods
Amato U;De Canditiis D
2001
Abstract
Wavelet and Fourier regularization methods are effective for the nonparametric regression problem. We prove that the loss function evaluated for the regularization parameter chosen through GCV or Mallows criteria is asymptotically equivalent in probability to its minimum over the regularization parameter. © 2001 Elsevier Science B.V.File in questo prodotto:
Non ci sono file associati a questo prodotto.
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
