1: \begin{abstract}

2: We consider the problem of estimating the predictive density of

3: future observations from a non-parametric regression model.

4: The~density estimators are evaluated under Kullback--Leibler divergence

5: and our focus is on establishing the exact asymptotics of minimax

6: risk in the case of Gaussian errors.  We derive the convergence rate

7: and constant for minimax risk among Bayesian predictive densities

8: under Gaussian priors and we show that this minimax risk is

9: asymptotically equivalent to that among all density estimators.

10: \end{abstract}