The numerical feasibility of a solution of Eq. (7) has been tested for a density estimation problem with one-dimensional and variables and the choice = , for which = . Choosing a true field (Fig. 1b), 50 data points have been sampled (Fig. 2a) from the corresponding true conditional likelihood = (Fig. 1a) with uniform . Including a Gaussian prior term for with zero mean and inverse covariance Eq. (7) has been solved on a mesh with 10 points in -direction and 15 points in -direction by standard iteration methods. (For technical details see [4].) The reconstructed conditional likelihood is shown in Fig. 2b, true and reconstructed regression functions are compared in Fig. 3.