Semi-supervised Learning

A form of ML combining supervised and unsupervised learning, using a small amount of labeled data and a large amount of unlabeled data for training.