Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

fastgae #6

Open
wagpa opened this issue Nov 23, 2022 · 2 comments
Open

fastgae #6

wagpa opened this issue Nov 23, 2022 · 2 comments
Labels
result This issue is or contains results to be used for the thesis source This issue contains or is a source used for the thesis

Comments

@wagpa
Copy link
Owner

wagpa commented Nov 23, 2022

https://github.com/deezer/fastgae

@wagpa wagpa added the source This issue contains or is a source used for the thesis label Nov 23, 2022
@wagpa
Copy link
Owner Author

wagpa commented Feb 1, 2023

Notes

  • "Graph autoencoders (AE) and variational autoencoders (VAE) are powerful node embedding methods, but suffer from scalability issues"
  • graph AE and VAE encodeing linear with number of edges per layer, decoding quadratic with number of nodes (+ memory issues for large n)
  • Proposes sampling method by importance. It results in a biased approximate loss
  • ...

@wagpa
Copy link
Owner Author

wagpa commented Feb 15, 2023

iterate over un-padded edges and calculate https://discord.com/channels/934839185855086662/988688735161946144/1070345614782632017

  • does the padding change the result?
  • weight edges higher than non-edges
  • use as loss function and metric

@wagpa wagpa added the result This issue is or contains results to be used for the thesis label Mar 31, 2023
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
result This issue is or contains results to be used for the thesis source This issue contains or is a source used for the thesis
Projects
None yet
Development

No branches or pull requests

1 participant