WSYCUHK_FDIA icon indicating copy to clipboard operation
WSYCUHK_FDIA copied to clipboard

Published 20 hours ago • verified publisher icon wsyCUHK

Questions About Dataset

Open sharkYLS opened this issue 3 years ago • 5 comments

Dear Dr. Wang, Hello, thank you for your great work. I have some questions about dataset. I would appreciate it if you could reply.

  • In y_test, y_train of your dataset, does 0 represent the bus without FDIA, and 1 represent the bus with FDIA?
  • If so, the sets of compromised data are far more than that of un-compromised data. But it is inconsistent with your README.md and paper. Is is rational?

@wsyCUHK

sharkYLS avatar Nov 30 '21 08:11 sharkYLS

Hi Shark, The 0 represent the bus without FDIA, and 1 represent the bus with FDIA. I'm sorry I may uploaded an old version of the data. I check the result, it might slightly different from the final results reported in the paper, due to the distribution changing. Actually, we later on find that we can share a better perforamce with 1:10 split and thus report the result with that. However, I'm really sorry again that the other data are not seeded.

Best, Shuoyao

wsyCUHK avatar Nov 30 '21 12:11 wsyCUHK

Thank you very much!

sharkYLS avatar Nov 30 '21 12:11 sharkYLS

Hi Shark, The 0 represent the bus without FDIA, and 1 represent the bus with FDIA. I'm sorry I may uploaded an old version of the data. I check the result, it might slightly different from the final results reported in the paper, due to the distribution changing. Actually, we later on find that we can share a better perforamce with 1:10 split and thus report the result with that. However, I'm really sorry again that the other data are not seeded.

Best, Shuoyao

Dear Dr. Wang, You mean that the data set in this repo is an old version, resulting in “does 0 represent the bus without FDIA, and 1 represent the bus with FDIA”? can you share the new version? Looking forward to your reply, it is very important to me.

ShuchunXu avatar Jul 08 '23 01:07 ShuchunXu

Hi Shark, The 0 represent the bus without FDIA, and 1 represent the bus with FDIA. I'm sorry I may uploaded an old version of the data. I check the result, it might slightly different from the final results reported in the paper, due to the distribution changing. Actually, we later on find that we can share a better perforamce with 1:10 split and thus report the result with that. However, I'm really sorry again that the other data are not seeded. Best, Shuoyao

Dear Dr. Wang, You mean that the data set in this repo is an old version, resulting in “does 0 represent the bus without FDIA, and 1 represent the bus with FDIA”? can you share the new version? Looking forward to your reply, it is very important to me.

I'm sorry but the other data is not seeded.

wsyCUHK avatar Jul 09 '23 04:07 wsyCUHK

Hi Shark, The 0 represent the bus without FDIA, and 1 represent the bus with FDIA. I'm sorry I may uploaded an old version of the data. I check the result, it might slightly different from the final results reported in the paper, due to the distribution changing. Actually, we later on find that we can share a better perforamce with 1:10 split and thus report the result with that. However, I'm really sorry again that the other data are not seeded. Best, Shuoyao

Dear Dr. Wang, You mean that the data set in this repo is an old version, resulting in “does 0 represent the bus without FDIA, and 1 represent the bus with FDIA”? can you share the new version? Looking forward to your reply, it is very important to me.

I'm sorry but the other data is not seeded.

Thank you very much!

ShuchunXu avatar Jul 09 '23 06:07 ShuchunXu