2
$\begingroup$

There is a popular strategy of using a neural network trained on one task to produce features for another related task by "chopping off" the top of the network and sewing the bottom onto some other modeling pipeline.

Word2Vec models employ this strategy, for example.

Is there an industry-popular term for this strategy? Are there any good resources that discuss its use in general terms?

$\endgroup$
1
$\begingroup$

This is an example of Transfer Learning, wherein instead of starting the learning process from scratch, a pre-trained model is used and appended with custom layers to achieve additional functionality.

enter image description here

| improve this answer | |
$\endgroup$
1
$\begingroup$

This a typical transfer learning technique, a lot of people refer to it with fine-tunning. I would recommend that you have a look on PyTorch tutorial: it explains well how to use.

| improve this answer | |
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.