We consider the problem of estimating an unknown function f from N noisy ob
servations on a random grid. In this paper we address the following aggrega
tion problem: given M functions f(1),...,f(M) find an "aggregated" estimato
r which approximates f nearly as well as the best convex combination f* of
f(1),...,f(M). We propose algorithms which provide approximations of f* wit
h expected L-2 accuracy O(N(-1/)4 ln(1/4) M). We show that this approximati
on rate cannot be significantly improved.
We discuss two specific applications: nonparametric prediction for a dynami
c system with output nonlinearity and reconstruction in the Jones-Barron cl
ass.