Estimação de Densidades de Dados Viesados Via Bases de Cossenos

Abstract

We consider first the problem of density estimation, we introduce a non-parametric estimator for this case, using a cosine basis. We apply the proposed methodology in a data set relative to the velocity of 82 galaxies measured in the region of Corona Borealis. Posteriorly, we presented the concept of biased data and we adapt the former estimator to this context. After that, we evaluate the results of this method using data simulated by an acceptance-rejection algorithm and later we apply this methodology in a sample taken from a dataset, containing information of blood alcohol concentration in drivers’ of the 50 american states, the District of Columbia, and Puerto Rico since 1975.