Hello, I have a question about training a neural network.
Assuming I have an already trained neural network, it would be possible to add one more entry to the trainningset without having to train all the neural network again?
If I understand you correctly then yes you can just add a new input-output pair i.e. question-answer to the training set, without retraining the network. However, why would you want to? As you will not be 'training' the network on the new pair, it wont make any difference.
If would make much more sense to add it to the test set and see how the network coped with an entry it has never seen before.
Does that make sense?
Log in to post a comment.
Sign up for the SourceForge newsletter:
You seem to have CSS turned off.
Please don't fill out this field.