Functional Derivatives in Statistics: Asymptotics and Robustness
摘要
Given a sample X 1, …, X n of i.i.d. random variables with common distribution function (df) F and empirical df F n , a statistic S( X 1, …, X n ) is called a statistical functional if it can be written in terms of a functional T, independent of n, such that S( X 1, …, X n ) = T( F n ), for all n ≥ 1. The domain of T contains at least the population df F and the empirical df F n , for all n ≥ 1. In this setting the statistic T( F n ) estimates the parameter T( F).