摘要：This paper deals with density and regression estimation problems for functional data.Using wavelet bases for Hilbert spaces of functions, we develop a new adaptive proce-dure based on a term-by-term selection of the wavelet co e.cients estimators.We study its asymptotic performances by considering the mean integrated squarederror over adapted decomposition spaces
关键词：functional data; density estimation; nonparametric regression; wavelets; hard thresh-;olding
