Empirical processes associated with V-statistics and a class of estimators under random censoring (Q1082745): Difference between revisions

Let \(h=h(X_{11},...,X_{1r_ 1},...,X_{k1},...,X_{kr_ k})\) be a kernel transform of degree \(r=(r_ 1,...,r_ k)\), taken from k-samples censored from the right. Denote with V the distribution function of h. In this paper, an estimate \(V_ N\) of V is considered, and an invariance principle for the corresponding empirical and quantile processes is derived. Applications to certain smooth functionals (such as L- statistics) are included. The paper constitutes an extension of existing results to the case of censored data.