Skip to content

Conversation

yupbank
Copy link

@yupbank yupbank commented May 9, 2019

Right now we sample both the neighbours and reversed neighbours. Following the paper, we only sample the reversed neighbours with rho ?

But the overall max_candidates is really interesting. i think it's a innovation from the code/practitioner

I think the origin paper missed a big detail, that the reversed neighbours would suffer from both too many and too little.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant