tsy935 / eeg-gnn-ssl Goto Github PK

Hello, I am replicating your code. The compressed package in the pretrained folder indicates that it is damaged. Can you upload it again? Thank you.

The roc_auc_score is used only in a binary classification context in this project.
The line if len(set(y)) <= 2: ensures that roc_auc_score is only called for binary classification.
In binary classification, multi_class is not required, hence avoiding the error.
However ， I run into a problem that raised error like this:

After debug , I found that the shape of y_prob_all in classification task is (103，4) and the shape of y_true_all is (103 , ).But the original code use if len(set(y)) <= 2: as condition , which would still result in error. I would like to ask for ur help

Mandarin:
在处理多分类任务时会遇到 raise ValueError("multi_class must be in ('ovo', 'ovr')")的问题，因为y的形状被调整为了（num，），无论如何len(set(y))的长度都是小于2的，多分类情况下y_prob_all形状为（num，classes）这会导致判断失效一直进入该条件而导致报错，请问你们的解决方法是什么？

Hi, I'm trying to reproduce all works in your paper. Could you share the code of other models (Dense-CNN, LSTM and CNN-LSTM)? Thanks!!

Great works!

I am reproducing your work and am having trouble visualizing the heat map (like Figure 3 and Figure 6 in the paper), can you please provide the code or describe the approach？

Thanks!

可以说明一下怎么跑通DOD-H数据集吗？

Hi siyi，

I am know reproducing your work, the work is very excellent, however, I have some questions to consult with you.

1. In folders "data", you have given pre-computed mean and standard deviation of the training data, however, I want to know the implementation details that could be applied in my data.
2. In fig.3, you used the Occlusion maps to visualize model result，could you please share the method in code.

I am looking forward to and appreciate your help!

Hello, sorry to bother you. Your code has taught me a lot. When I ran your densecnn data, I found that the program reported the following error, which means that resample_freq does not exist.
（您好，打扰您了，您的代码让我学习到了很多东西，我在运行您的densecnn数据的时候，发现程序报错如下，意思指的是不存在resample_freq。）

In the part of your data preprocessing, I noticed that this resample_freq is outside the with. I would like to ask if this line of code should be inside the ‘with‘? Thank you for your answer, I wish you a happy life~
（在您的数据预处理部分，我注意到这个resample_freq是在with外面，想请问一下这行代码应该是在with里面吗？谢谢您的解答，祝您生活愉快~）

Great work!

I'm trying to reproduce this excellent work. I get some trouble during data preprocessing. For the data resampling, I have the following error:

ValueError: 'EEG FZ' is not in list.

Could you please provide some suggestions about this?

Thank you!