Skip to content

About the min-max modeling. #3

Description

@cjfcsjt

The research work is exciting! I noticed that you formulate the automated graph contrastive learning as the min-max optimization problem. May I ask that why not use the bi-level optimization where the upper level objective is to search the augmentation operations and the lower level objective is to optimize the contrastive loss (like existing one-shot neural architecture search methods)?

Activity

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

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions