# Self-training (machine learning)

> A variant of self-supervised learning that is particularly useful when all of the following conditions are true:[^1]

A variant of <self-supervised learning> that is particularly useful when all of the following conditions are true:[^1]

- The ratio of [unlabeled examples](https://wiki.g15e.com/pages/Unlabeled%20example.txt) to [labeled examples](https://wiki.g15e.com/pages/Labeled%20example.txt) in the dataset is high.
- This is a [classification model](https://wiki.g15e.com/pages/Classification%20model.txt) problem.

Self-training works by iterating over the following two steps until the model stops improving:

1. Use [supervised learning](https://wiki.g15e.com/pages/Supervised%20learning.txt) to train a model on the labeled examples.
2. Use the model created in Step 1 to generate predictions (labels) on the unlabeled examples, moving those in which there is high confidence into the labeled examples with the predicted label.

Notice that each iteration of Step 2 adds more labeled examples for Step 1 to train on.

## Footnotes

[^1]: https://developers.google.com/machine-learning/glossary#self-training