{% extends "Test-layout.html" %} {% block content %}
Semi-supervised learning is a combination of supervised and unsupervised learning. It uses a small amount of labeled data and a large amount of unlabeled data to train a model. This is particularly useful when it is expensive to label data.