Training ML models on unlabeled data, where the model identifies patterns and relationships within the data itself.
