Workshop Learning Objectives (Transfer Learning)

1. Operationally define transfer learning.
2. Describe the purpose of transfer learning.
3. Discuss the advantages of transfer learning.
4. Discuss the three key questions to ask related to transfer learning.
5. Discuss the different types of transfer learning.
6. Discuss how layers (deep learning) can be cut or switched in terms of transfer learning.
7. Implement classification with transfer learning.
8. Discuss VGG16 model and ImageNet.
9. Describe how a new deep neural network (DNN) could be done by reusing layers.
10. Discuss what is Hugging Face.
11. Implement fine tuning with transfer learning.
12. Discuss the importance of freezing and fine-tuning certain layers.

---

See Deep Learning for Coders with fastai and Pytorch, page 31-ff for discussion of pretrained models, transfer learning, and the absence of content on this subject.