Asymptotic normality of adaptive wavelet thresholding risk estimation (Q1761079)

Wavelet decomposition of a signal can be used for the denoising of signals/images. The process of using threshold value for discarding/retaining the wavelet coefficients depends upon the problem and the purpose and may involve error/risk. The author studies the asymptotic normality of risk estimation for the one dimensional signals for the adaptive threshold which has been suggested by \textit{D. Donoho} and \textit{I. M. Johnstone} [J. Am. Stat. Assoc. 90, No. 432, 1200--1224 (1995; Zbl 0869.62024)].