In the context of cross-validation for regularized regressio…
In the context of cross-validation for regularized regression, the prediction risk estimate for a given value of lambda is obtained by averaging the scoring values across all folds of the data.