Mehta, A. (2019, June 29). Retrieved from [www.digitalvidya.com/blog/semi-supervised-learning/](https://www.digitalvidya.com/blog/semi-supervised-learning/) 

University of Michigan - Ann Arbor

This image shows the pseudo labeling process:
1. Labeled data is used to train a model
2. Unlabeled data is then labeled by the trained model (pseudo labeled data)
3. Both the pseudo labeled data and labeled data are then combined into the full dataset
4. The full dataset is then used to retrain the model


Learn Before

Related