At cifar10 0.2 sym noise rate setting, cann't repeat the results reported at paper.

Open sillytheif opened this issue 3 years ago • 4 comments

The cifar10 0.2 noise rate, I can only get about 91.5% acc useing the code, and I can't find the reason

Apr 19 '23 03:04 sillytheif

I found a similar question and got the same precision as yours.

May 23 '23 08:05 kovelxyz

I can only get ~92% performance as well.

Jul 01 '23 23:07 xhl-video

I have found the solution. We should follow Table 7 in the appendix to assign λu. It should be 0 for the Cifar10 20% dataset, which is pretty different for other noise ratios.

Jul 06 '23 01:07 xhl-video