Dataset class name?

[Aceticia]

The datas is provided but I can't seem to find which label is which class. Can you provide the order of the 6 labels in a list?

The training of your model is done using all_train_X.npy and all_train_Y2_hard.npy, but the label file's class counts don't match with either dataset 1 or 2 in your paper.

[MohSamNaf]

I can possibly assume it is SZ, LPD, GPD, LRDA, GRDA and OTH. Whenever, they mention the classes, the class types are always mentioned into this order.
Additionally, I noticed there is no file for the labels of the test dataset (Dataset 4). Did you figure this out?

[Aceticia]

I'm not sure if we can assume this since some classes are mentioned in different orders in the paper.

Additionally, I noticed there is no file for the labels of the test dataset (Dataset 4). Did you figure this out?

I believe this is done intentionally so you can't train on test set. If you want to evaluate, you will need to send your model and evaluation script to the authors. However, given the lack of responses to this repo, I think it's unlikely that they will provide evaluations. But maybe you should still try.

[MohSamNaf]

I'm not sure if we can assume this since some classes are mentioned in different orders in the paper.

Additionally, I noticed there is no file for the labels of the test dataset (Dataset 4). Did you figure this out?

I believe this is done intentionally so you can't train on test set. If you want to evaluate, you will need to send your model and evaluation script to the authors. However, given the lack of responses to this repo, I think it's unlikely that they will provide evaluations. But maybe you should still try.

May you tell me the paper title that they mention the different order?

Edit: I believe you are right. I conducted a little analysis on the 10_train_Y2_hard.npy file. I counted the 1s across each index.
The output was [ 6909 545 7578 10040 6077 5093] which corresponds to [Other, Seizure, LPD, GPD, LRDA, GRDA]. I matched these values with the labels from Figure S2 in the Supplemental material for Training Dataset 3.

import numpy as np

path = 'J:\SparcNet Dataset\\10_train_Y2_hard.npy'
data = np.load(path)
ones_count = np.sum(data == 1, axis=0)
print(ones_count)