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

Issue about metrics #1

Open
night3759 opened this issue Jan 4, 2023 · 2 comments
Open

Issue about metrics #1

night3759 opened this issue Jan 4, 2023 · 2 comments

Comments

@night3759
Copy link

Thank you for your job! I got a question when I was reading the code, I found there is a mask operation during calculating the MAE and the MAPE. I don't know why there is a mask, and why the mask value is 5. I am looking forwark to your reply. Thank you.

@Echo-Ji
Copy link
Owner

Echo-Ji commented Jan 9, 2023

Thanks for your comments.

  • Why add a mask to MAE?
    The goal is to filter samples with small values that may be caused by noise or error (such as a loop sensor).

  • Why add a mask to MAPE?
    The goal is to ignore very small values (e.g., 0.5/0.5=100%).

The mask operation is borrowed from AGCRN [1], and 5 is a common practice used in industry and academia.

[1] Bai, L.; Yao, L.; Li, C.; Wang, X.; and Wang, C. 2020. Adaptive graph convolutional recurrent network for traffic forecasting. NeurIPS, 33: 17804–17815.

@liuyixun
Copy link

如果这么做的话,mask_value以下的值就预测不到了,相当于完全放弃了mask_value以下的值了?

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