Graph Contrastive Learning with Augmentations笔记

NeurIPS 2020- Graph Contrastive Learning with Augmentations

• • contrastive learning algorithm
  • pretraining model for molecular proporty predition

使用最基础的contrastive loss 处理图graph-level的tasks, 包括self-supervised, semi-supervised graph classification,

主要贡献是提出4种不同的augmentations.

contrastive learning algorithm

使用最经典的方法

pretraining model for molecular proporty predition

本文发现不同数据集适合不同的data 增广的方式