What is semi-supervised learning?

Semi-supervised learning is a machine learning paradigm that uses both labeled and unlabeled data for training. It is particularly useful when labeled data is scarce or expensive to obtain, as it leverages the readily available unlabeled data to improve the performance of the model. Common applications include image classification, text classification, and speech recognition.