Abstract
Under the assumption that the distribution function of the observation noise is known, both for the case of a predefined observation design and the case where observation designing is possible, we construct estimates of smooth functionals of the regression function, for which lower bounds on mean-square risks of arbitrary estimates of smooth functionals obtained in [1, 2] are asymptotically attained.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have