Closed rickyang1114 closed 1 year ago
Hi, actually we mentioned in page 4 that we use all other features in the mini-batch as B_i, and also explained that there may exist the situation that B_i has intersection with C_i (we think it is fine, as it should not be frequent).
However, you can try it with excluding the potential neighbors in B_i, it may improve the performance.
My doubts are clarified. Thanks for your reply.
Dear author,
I'm an undergraduate student who is quite interested in SFDA, and I have been following your work from G-SFDA, NRC to AaD. I really appreciate your works and the wonderful performance they achieve.
Recently, I'm trying to understand your work of AaD. However, I become a little confused about some implementation details as I conbine your paper and source code.
From my understanding, the
B_i
in div term contains all other items in a mini batch except those inC_i
. In other words, items that are k nearest neighbors should be excluded fromB_i
, as presented in the paper. However, in your code I copied below:it seems that only diagonal entries in
mask
are set to 0, rather than k nearest neighbors.I suppose I must have some misunderstandings, so I create the issue, hoping to get your answer. I would appreciate it if you could answer my qusetion. Looking forward to your reply.