Which of the following statements about cross-validation is true?

The statement about k-fold cross-validation requiring fitting the model k times is accurate. In k-fold cross-validation, the dataset is divided into k subsets or "folds." The model is trained on k-1 folds while using the remaining fold for validation. This process is repeated k times, with each fold being used as the validation set once. Therefore, the model must be fitted a total of k times, which directly aligns with the correct interpretation of how k-fold cross-validation operates.

Regarding the other options, the concept of Leave-One-Out Cross-Validation (LOOCV) involves fitting the model for each observation in the dataset, which means that the number of fit operations equals the number of observations, not fitting once for the entire dataset. The notion that LOOCV can be efficient with small datasets is theoretically true to an extent but becomes less efficient in larger datasets due to the excessive number of fits required. Lastly, cross-validation techniques are definitely applicable to both regression and classification models, which invalidates the statement suggesting otherwise.