Semi-supervised Learning: A type of machine learning where the model is trained on a labeled dataset but also uses unlabeled data to improve its performance.