aprendizaje semisupervisado

Also known as weak supervision

machine learning approach where noisy, limited, or imprecise sources are used to provide supervision signal for labeling large amounts of training data in a supervised learning setting