Is normal data supposed to have non-zero attack values?

[rightx2]

Looks like all abnormal datasets have non-zero attack values in the period, but normal datasets don't.

• "normal_20190911T200000_to_20190915T100000.csv": has all zero values
• "normal_20191101T200000_to_20191104T150000.csv": has some non-zero values

attack       776
attack.P1    776
attack.P2      0
attack.P3      0
dtype: int64

I thought that the attack values made the data abnormal.. Did I miss something?

[wml20]

You are right.
I checked the normal data and confirmed that SA2 (P1-PCPSP1PV1) attacks occurred periodically on November 2nd at 15:31 and November 3rd at 15:31.

This is caused by the operation of a part of the attack program during the normal data collection period.

We will update the normal data that deleted the attack part soon.

Thank you.

[mzietlow]

Any updates on this?

Best wishes
~

[dolgam]

We will upload the modified version of HAI 1.0 until 17th Feb. (--> It will be postponed to March. )
Sorry for the late update.
In addition, a new HAI dataset will be opened on the same day.

[mzietlow]

Oh, the postponement is really unfortunate because I wanted to use the data in my thesis and need it by early March at the latest. Thank you for keeping us updated

[dolgam]

I am sorry about the postponement due to our internal checking process.
We are trying to finish it early.

[yif0]

Any updates on this? Thanks for your effors!

[dolgam]

We fixed a few bugs (not critical) of HAI 1.0 and made a new HAI dataset. After the internal process, we will upload them.