Estimates of distributions of components in a mixture from censoring data
Anton Ryzhov · Theory of Probability and Mathematical Statistics · 2005
The problem of estimation of the distribution functions of components in a mixture in the case of censored observations is considered. Optimal estimators are found in the class of linear estimators. Since the optimal estimators depend on unknown distribution functions of components, an adaptive estimation scheme is used. The asymptotic normality is proved for adaptive estimators and it is shown that their concentration coefficient coincides with that of the optimal linear estimator.