0
$\begingroup$

I have a multi-label text classification dataset which is very small around 80Kb, I am only going to receive a small amount of data for training from my client. But it is expected to build a high accuracy model for classifying each utterance into multiple targets based on the utterance. How should i prepare the dataset, which models would be best and how shall I aim for the highest accuracy

$\endgroup$

1 Answer 1

1
$\begingroup$

You could use a pre-trained model like Bert for feature extraction. Then use these features to train your own classifier using the data received.

$\endgroup$
1
  • $\begingroup$ thanks for the advice! $\endgroup$ Feb 28 at 19:22

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