Inconsistent result of BGRL on Wiki-CS dataset
ZsZsZs25 opened this issue · comments
Thanks for your great work fiirst! But the experiment result of BGRL on Wiki-CS dataset of your paper is much inconsistent with the BGRL paper and your own experiment result in repository BGRL_Pytorch. Did I make any stupid mistakes?
Thanks for your interest to our paper.
For WikiCS, in preprocessing step, we only implemented feature-wise normalization, while BGRL did feature-wise normalization and data normalization also. They also only implemented feature-wise normalization on other datasets.
We compared our model and all other models under the same condition (only feature-wise normalization).
For more detail on BGRL, please refer to supplementary material of their submission at ICLR 2022 (https://openreview.net/forum?id=0UXT6PpRpW)
I fixed my BGRL_Pytorch code after openreview opened (sep 29 2021) .
Thanks!