Breaking the curse of dimensionality for linear rules: optimal predictors over the ellipsoid

Open in new window