What is ‘semi-supervised learning’?

A A learning paradigm using a small amount of labeled data combined with a large amount of unlabeled data for training B Supervised learning with only half the training epochs C Learning with a partially trained model on full data D A method combining two supervised algorithms