A note on the wavelet deconvolution of a density from mixtures under quadrant dependence

Christophe Chesneau (1)

1 LMNO - Laboratoire de Mathématiques Nicolas Oresme

We consider the density convolution model: $Y=X+\epsilon$, where $X$ and $\epsilon$ are independent random variables. We suppose that the density of $X$ is a finite mixture with unknown components. We want to estimate a component of this mixture from pairwise positive quadrant dependent observations $Y_1,\ldots,Y_n$. To reach this goal, a linear wavelet estimator is developed. We measure its performance by determining an upper bound of the mean integrated squared error.

Mots clés: Density deconvolution; Mixture; Quadrant dependence; Wavelets; Besov balls

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

Christophe Chesneau. A note on the wavelet deconvolution of a density from mixtures under quadrant dependence. 2010. ⟨hal-00518416⟩