search

eeyhsong / EEG-Conformer

EEG Transformer 2.0. i. Convolutional Transformer for EEG Decoding. ii. Novel visualization - Class Activation Topography.
GNU General Public License v3.0
389 stars 55 forks source link

accuracy lower than that in the paper #30

Open YamengGu opened 4 months ago

YamengGu commented 4 months ago

Hi,

I edited the input data from dataset 2a which included both data and label. I run the code and found that each subject's accuracy is lower than that you described in the paper. What should I do? for example, my subject 1 best accuracy is 77% while yours is 88. my second subject is 47% while yours is 61%. other subjects keep the same trend.

Here is the link for my input data, just in case my input data is wrong: https://drive.google.com/file/d/1WQaB4G2i6z8r6drgbJ9gF5aZCrA_Ydek/view?usp=sharing

Thanks,

Yameng

eeyhsong commented 4 months ago

Do you use the example data I supported in the google driver?

YamengGu commented 4 months ago

No. I used the data i created in the shared Google drive. It is very similar to the 01E data you shared.Best,YamengSent from my iPhoneOn Apr 14, 2024, at 10:34 AM, eeyhsong @.***> wrote: Do you use the example data I supported in the google driver?

—Reply to this email directly, view it on GitHub, or unsubscribe.You are receiving this because you authored the thread.Message ID: @.***>

eeyhsong commented 4 months ago

Where did you get the data? Did you perform some preprocessing such as band-pass filter and z-score? refer to code.

Best, Song