Leave-One-Out Cross-Validation: A technique used to evaluate the performance of a model by training and evaluating the model n times, where n is the number of samples in the data, and each time leaving out one sample from the training set.