Nonparametric estimation of smooth probability densities in \(L_ 2\) (Q1204666)

Let \(x_ 1,x_ 2,\dots,x_ n\) be a sequence of i.i.d. r.v., having unknown density \(p(x)\), \(x \in \mathbb{R}\). The author constructs asymptotically minimax estimators of \(p(x)\) based on the sample \(x_ 1,\dots,x_ n\). The quality of the estimation is measured by a quadratic risk function. A priori information on the smoothness of \(p(x)\) is absent.