Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Number of classes for age classification #7

Open
erobic opened this issue Jul 1, 2020 · 3 comments
Open

Number of classes for age classification #7

erobic opened this issue Jul 1, 2020 · 3 comments

Comments

@erobic
Copy link

erobic commented Jul 1, 2020

How many output units do you use for age classification? Is each age a separate class? Or is it a binary problem 0-29 vs 40+?

@zwvews
Copy link

zwvews commented Sep 6, 2020

Hi, I think they should use the binary label since I found that the number of testing data stated by their paper is 13129. This is exactly the sum of data with age <=29 and >40

@erobic
Copy link
Author

erobic commented Sep 8, 2020

Hi,

I asked the authors about this and they replied that they use 12 nodes for age classification with intervals from https://arxiv.org/abs/1809.02169. Unfortunately, I wasn't able to replicate the numbers with that setting though. Even the baseline numbers were off, so I asked for hyperparameters/other settings, but haven't heard back since then.

Also, treating this as binary classification problem would lose the distinction between age classification vs gender classification, so I think # of classes has to be different than 2.

@carloalbertobarbano
Copy link

I agree with @erobic.
@feidfoe could you provide us with more info? thanks

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

3 participants