Cross Validation is a model validation technique in which the available dataset is split into various subsets, or "folds", and each fold is used once as a validation set, with the complement being used as the training set. For example: in 10-fold cross validation, the dataset is split in 10 parts, and at each of 10 iterations 1/10th of the data is used for validation, and 9/10ths for training. In a dataset with N elements, a N-fold cross validation is more commonly known as a "leave-one-out" cross validation.