-
Notifications
You must be signed in to change notification settings - Fork 453
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
第五章代码有错误 #54
Comments
GCN里的划分就是140:500:1000,每个label取20个做训练集,不要想当然啊 |
为什么训练集节点这么少啊,验证和测试的却这么多
发自我的iPhone
…------------------ 原始邮件 ------------------
发件人: oldppd ***@***.***>
发送时间: 2021年4月20日 21:50
收件人: FighterLYL/GraphNeuralNetwork ***@***.***>
抄送: Ireliafans ***@***.***>, Author ***@***.***>
主题: 回复:[FighterLYL/GraphNeuralNetwork] 第五章代码有错误 (#54)
GCN里的划分就是140:500:1000,每个label取20个做训练集,不要想当然啊
—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub, or unsubscribe.
|
3个citation graph上用这个划分很早就有了,是每个label取20个作为训练集,cora是7类,所以就是140个作为训练集 |
您好我想问下这个分类有什么依据在吗 |
从Kipf的GCN [1]开始,Cora、Citeseer、Pubmed这老三样citation graph就一直用这个划分了,每个类别20个作为训练集 [1] Thomas N. Kipf and Max Welling. 2017. Semi-Supervised Classification with Graph Convolutional Networks. In International Conference on Learning Representations (ICLR). |
太牛逼了,谢谢您!终于找到这是为啥了 |
第五章的数据有2708个样本,但是训练集验证集 测试集 数据分布分别是140:500:1000,数据划分存在问题。
正确的数据集划分应该把
“train_index = np.arange(y.shape[0]) 数据集大小有问题
val_index = np.arange(y.shape[0], y.shape[0] + 500)”
修改为
“x_shape=1208
train_index = np.arange(x_shape)
val_index = np.arange(x_shape, x_shape + 500)”
数据集分布变成了1208:500:1000 ,测试集效果能从0.805 提高到0.871
The text was updated successfully, but these errors were encountered: