Skip to content
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

论文中table8中的数据是怎么得到的呢 #2

Open
qiljj opened this issue Feb 5, 2023 · 8 comments
Open

论文中table8中的数据是怎么得到的呢 #2

qiljj opened this issue Feb 5, 2023 · 8 comments

Comments

@qiljj
Copy link

qiljj commented Feb 5, 2023

运行f3mend出现以下问题:
b6a44842c1cb7a6da07995c0012ef77
我把nn.liner那里的参数用int()强制转换也出错
3c866d8ff203550d6db20220d01379e
b6a44842c1cb7a6da07995c0012ef77
nn.linear接受的应该是int类型的参数,但是mlpdim还有lnndim这两个好像它识别的是str类型的,所以导致出现了错误(这个是我自己的理解)

@easezyc
Copy link
Collaborator

easezyc commented Feb 6, 2023

我确认在我这跑没有问题,不太清楚为什么你的机器上会出这个问题。

@qiljj
Copy link
Author

qiljj commented Feb 6, 2023 via email

@qiljj qiljj changed the title 数据类型问题 论文中table8中的数据是怎么得到的呢 Mar 10, 2023
@qiljj
Copy link
Author

qiljj commented Mar 10, 2023

前辈,我仔细研读了您的该论文,我想请问您我的理解对吗?table8中前9列平均F1值=通过连续跑10次然后取各个领域的F1-score的平均值。第二个问题就是table8 后三列(overall)F1 Acc AUC数值是怎么计算得到的呢?
image

@qiljj
Copy link
Author

qiljj commented Mar 10, 2023

补充一个问题:在消融实验中,您提到了table10 ,在论文中显示是总体F1评分,我也不太明白这个数值是怎么得到的,方便讲解一下吗?
image

@easezyc
Copy link
Collaborator

easezyc commented Mar 10, 2023

overall不是前面几列取平均,而是样本级别的平均(micro),简单说就是不区分领域标签,所有样本直接算指标。

@qiljj
Copy link
Author

qiljj commented Apr 4, 2023

好,已明白。您在文中提到:We record average results over ten runs. We report accuracy (Acc), macro F1 score (F1), and Area Under ROC (AUC).所有的数据您是连续跑了十次,每跑一次不是得到一个最优模型嘛,那您的意思是跑10次,得到10个最优模型和10个best metric,你是将跑10次的best metric取平均了吗?(还是说就取一次best metric就行) 这个问题有点困扰我,希望得到您的回复

@qiljj
Copy link
Author

qiljj commented Apr 4, 2023

上面问题的意思就是:您是连续跑了10次同一个模型吗?or 只跑一次,取最好的结果

@easezyc
Copy link
Collaborator

easezyc commented Apr 6, 2023

跑十次平均

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants