STK_PARAM_LOOMSE computes the Leave-One-Out criterion of a model given data CALL: [C, COVPARAM_DIFF, LNV_DIFF] = stk_param_loomse (MODEL, XI, ZI) computes the value C of the leave-one-out mean square error criterion for the MODEL given the data (XI, ZI). CALL: [C, COVPARAM_DIFF, LNV_DIFF] = stk_param_loomse (MODEL, XI, ZI) also returns the gradient COVPARAM_DIFF of C with respect to the parameters of the covariance function, and its derivative LNV_DIFF with respect to the logarithm of the noise variance. REFERENCE [1] Francois Bachoc. Estimation parametrique de la fonction de covariance dans le modele de krigeage par processus gaussiens: application a la quantification des incertitudes en simulation numerique. PhD thesis, Paris 7, 2013. http://www.theses.fr/2013PA077111 See also: stk_predict_leaveoneout, stk_param_relik