Giter Club home page Giter Club logo

merit's Introduction

MERIT

A PyTorch implementation of our IJCAI-21 paper Multi-Scale Contrastive Siamese Networks for Self-Supervised Graph Representation Learning.

Dependencies

  • Python (>=3.6)
  • PyTorch (>=1.7.1)
  • NumPy (>=1.19.2)
  • Scikit-Learn (>=0.24.1)
  • Scipy (>=1.6.1)
  • Networkx (>=2.5)

To install all dependencies:

pip install -r requirements.txt

Usage

Here we provide the implementation of MERIT along with Cora and Citeseer dataset.

  • To train and evaluate on Cora:
python run_cora.py
  • To train and evaluate on Citeseer:
python run_citeseer.py

Citation

If you use our code in your research, please cite the following article:

@inproceedings{Jin2021MultiScaleCS,
  title={Multi-Scale Contrastive Siamese Networks for Self-Supervised Graph Representation Learning},
  author={Ming Jin and Yizhen Zheng and Yuan-Fang Li and Chen Gong and Chuan Zhou and Shirui Pan},
  booktitle={The 30th International Joint Conference on Artificial Intelligence (IJCAI)},
  year={2021}
}

merit's People

Contributors

dependabot[bot] avatar kimmeen avatar

Stargazers

avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar

Watchers

avatar

Forkers

trendingtechnology sean0719 hanbei969 abram07 womenhenzhiliu lyndonlens s3cl-code 2016102050016

merit's Issues

Questions about accuracy.

According to the implementation of other codes such as DGI, GRACE, etc., they all use the results of the last epoch or the minimum loss in the training process (early stopping) to calculate acc as the result. And your code is to calculate acc every 10 rounds, and use the best acc among all acc as your experimental result.

I think contrastive learning is unsupervised learning, and their codes are correct. It is unfair to compare the acc of your codes with them.

Why is the acc different from the MVGRL paper?

Hello! In your paper, your acc of MVGRL method on three datasets (Cora, CiteSeer, PubMed) are 82.9, 72.6, 79.4. But in MVGRL paper, the acc are 86.8, 73.3, 80.1. I think both of you use the same dataset, so could you please explain the reason?

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.