Search Results

Semi-supervised learning

Semi-supervised learning mixes labeled and labeled data to produce better models.

In machine learning, finding or creating correctly-labeled data (e.g. images with annotations and descriptions, audio with text transcriptions) can be difficult or expensive.

Semi-supervised learning techniques take advantage of a small amount of labeled data and a large amount of unlabeled data to produce a better model than a purely supervised learning or a purely unsupervised learning technique.