3. Transfer Learning

Transfer learning corresponds to a way to reuse information obtained by previous model training for a second related task. This can be necessary when only a concise amount of data is available. Therefore, a PyTorch based Jupyter Notebook provides the implementation of Transfer Learning. This technique was illustrated for the same setting as the first example, predicting missing land surface temperature values.

Demo Notebook

• Transfer Learning

The Basic Concept of Transfer Learning.