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.

Sign up for GitHub

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

Jump to bottom

Why normalization and resample ? #10

Open

zhgqcn opened this issue Apr 4, 2022 · 2 comments

zhgqcn commented Apr 4, 2022 •

edited

Loading

Dear Mr.Zhong, may I ask some questions aboult the code.

Why resample the dataset in the preprocessing stage, the uk-dale dataset original were sampled by 6 second, and I don't know why resample.

I attempt to read the paper the find the answer but don't get the useful ideas.
My guess is to have the same experimental conditions as the previous paper.

Why normalize the dataset, after normalizing, the target is so closer that I think the net can't distinguish them well?
How to choose the suitable time to test, I read the test.csv and find most time of the appliance is in no work condition.

Looking forward to your answer, thank you very much.

qinxiaopang commented Apr 7, 2024

Have you firgured it out?

Owner

MingjunZhong commented Apr 8, 2024

If I remember correctly, resampling was to make the input mains reading and targets to have the same time stamps. It also makes to fill NaN entries if needed.
Normalizing has been recognised as a key technique to help the model to converge faster in general; this is also true for this particular problem.
To test the model, you should randomly choose a long time series.
Hopefully this would help you to understand the problems.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment