Is there any difference between predictAndUpdateState function and transfer learning?

Hellow,

I was looking at the functionality of LSTM and found predictAndUpdateState.

As MATLAB describes, it keeps the previously trained network and updates it with new data.

This is the same process that transfer learning learns from new data. Are these two functions the same?

First of all, thank you.

Answer 1

[Prateek Rai]

To my understanding, you want to know the difference between predictAndUpdateState and Transfer Learning. "predictAndUpdateState" is a function that predicts responses using a trained recurrent neural network and updates the network state, whereas, "transfer learning" is a process where you can take a pre-trained network and use it as a starting point to learn a new task.

You can refer to predictAndUpdateState MathWorks documentation page to learn more on predictAndUpdateState. You can also refer transfer learning using alexnet to MathWorks documentation page to learn more on transfer learning.