That supervised learning is learning from examples, where the training dataset is separate from the test dataset.
That unsupervised learning is when there is no labeled training dataset, you are learning from data without explicit labels.
That semi-supervised learning is a mixture of unsupervised learning and supervised learning, where the training dataset is a mix of labeled and unlabeled examples.

#machine-learning #unsupervised-learning #supervised-learning #data-science #data-analysis