Estimation and Prediction of Nonlinear Functionals of Gaussian Processes.

摘要

The general estimation problem of nonlinear functionals of Gaussian processes is solved via the tensor product structure of nonlinear space in the sense that the nonlinear problem reduces to a standard linear estimation problem, the theory of which has been well developed. Also introduced the concept of super predictor for a class of prediction problems and a lower bound for the mean square error of the nonlinear prediction is derived.