I was wondering if someone could clearly explain the extent of transfer learning, fine-tuning, and domain adaptation. From my understanding, both fine-tuning and domain adaptation are subcategories of transfer learning. I am not very clear about the difference between domain adaptation and fine-tuning.
1 Answer
I also found difficult to find a clear definition of those terms. One of the best references for that matter is: Sinno Jialin Pan and Qiang Yang. A Survey on Transfer Learning. IEEE Transactions on Knowledge and Data Engineering, 2009
Here is a very nice diagram showing you the differences. As you point out, they are names given to particular use cases of transfer learning. Depending on how tuning is done, for example unsupervised training on some task and then fine tunning on the target task (inductive transfer learning), it can fall in one category or the other.