Semi-supervised Learning
Last Updated: July 29, 2026 | By Mihail Sebastian | AI Dictionary
A machine learning approach that combines a small set of labeled examples with a large pool of unlabeled data, useful when labeling is expensive.
What is Semi-supervised Learning?
Semi-supervised learning is a machine learning approach that trains a model on a small set of labeled examples combined with a much larger pool of unlabeled data.
It sits between supervised learning, which needs every example labeled, and unsupervised learning, which uses no labels at all. The approach pays off when labels are expensive to produce but raw data is abundant.
How Semi-supervised Learning Works
Training starts with the labeled set: the model learns the basic input-to-output mapping from a few hundred or a few thousand answered examples. It then turns to the unlabeled pool. In self-training, the most common technique, the model labels the unlabeled examples itself, keeps only the predictions it is most confident about, and retrains on the enlarged dataset.
The unlabeled data helps because it reveals the shape of the data: where examples cluster and where the natural boundaries between groups fall. The few real labels anchor those clusters to actual categories.
Example of Semi-supervised Learning
A hospital wants a model that flags possible tumors in X-rays. Radiologists have annotated 500 images, which is labeling at its most expensive – expert hours per image – while the archive holds 100,000 unannotated scans.
The team trains a first model on the 500 labeled images, runs it over the archive, keeps only its high-confidence predictions as new labels, and retrains on the larger set. Accuracy improves without another radiologist hour, because the archive taught the model what X-rays look like in general before it had to decide what a tumor looks like.
Related AI terms: Supervised Learning · Unsupervised Learning · Machine Learning · Labeling · Few-shot Learning
Did you like the Semi-supervised Learning gist?
Learn about 250+ need-to-know artificial intelligence terms in the AI Dictionary.
Mihail Sebastian — Writes about AI governance, regulation, and the technology behind them. Placeholder bio — replace with a real credential line. About