3
$\begingroup$

In a speaker recognition problem I have 330 speakers (classes) as targets and want to predict the identities with a feedforward neural net with a softmax output layer.

The thing is some classes have as much as 10 times more training data available compared to the smaller classes. If I don't do any balancing of the classes the results are kind of bad cause the big classes tend to dominate and the smaller classes are often misclassified.

One easy solution I can use is to throw away lots of training data for the bigger classes in order to have a balanced dataset and speed up the training a bit. It kind of works but it seems to be very sub-optimal.

Maybe I could try replicating some samples from the smaller classes to have as much as the bigger class? This would lead to slower training but at least I wouldn´t throw away real training data.

I was wondering, is there a more elegant way of weighting the importance of the classes during training or something like that? In order to get a better accuracy without having the bigger classes "eating" all the small classes? It would be great if there was a method to do so without having to upsample or downsample the training set of individual classes, but keeping an unbalanced set as an input.

$\endgroup$
2
$\begingroup$

You can use different weights for different samples or categories when computing the cost during training, (say using a higher cost for uncommon samples/categories than the others).

Such is called the cost-sensitive method according to this paper, and there're also many other methods mentioned in it. I first found the paper here in a similar question.

There's also an question (unanswered at the time of writing) Tuning priors/weights/costs to counteract class imbalance that might be related.

| cite | improve this answer | |
$\endgroup$
  • 1
    $\begingroup$ This very recent paper seems to claim that weights do nothing for deep neural networks (I didn't read it carefully yet) arxiv.org/abs/1812.03372 $\endgroup$ – Tim Dec 12 '18 at 11:56
  • $\begingroup$ Me neither, the paper appeared on arXiv four days ago :) $\endgroup$ – Tim Dec 12 '18 at 13:38

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.