[neaolei1]

Dear Organizers,
after analysis, we find that in the training set there are some parts of pitting level 2 whose amplitude is close to or even higher than the pitting level 8, which is not in line with common sense. Therefore we have doubts about whether there are cases of data labeling errors, or whether this situation is normal and there are such samples in the test set.

I am appreciate for your reply.
NeaoLei

[yongzhiqu]

Hello Neaolei,

Thank you for the message.

The amplitude is generally in line with the fault, however that is not always the case. In some or many operation conditions, there are cases that a small fault may have a higher amplitude. Similar pattern will appear in the testing and validation as well.

The data labels are all correct.

Best Regards,
PHM Data Challenge Team.

• This reply was modified 9 months, 2 weeks ago by yongzhiqu.

[Author]

Posts