Bootstrapped Representation Learning on Graphs

By Shantanu Thakoor, Corentin Tallec, Mohammad Gheshlaghi Azar, Mehdi Azabou, Eva L. Dyer, R´emi Munos, Petar Velivckovi'c, Michal Valko. arXiv.org, 2021

315 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/58d3278e45284e2910d5d972b2ec2c7ce6e34c4f

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

1. Weihua Hu
2. Shirui Pan
3. Yixin Liu
4. Yuxiao Dong
5. Matthias Fey
6. Philip S. Yu
7. Jure Leskovec
8. Feng Xia

Citations in the corpus

• Graph Self-Supervised Learning: A Survey (2021, methodology): “Biochemical Graphs PPI [137] 24 56,944 818,716 50 121* [13], [33], [41], [49], [57], [77], [78] [81], [84], [87]”
• Self-Supervised Learning on Graphs: Contrastive, Generative, or Predictive (2021, methodology): “SelfGNN [35] differs from BGRL only in the definition of the objective function.”
• From Canonical Correlation Analysis to Self-supervised Graph Neural Networks (2021, methodology): “In this subsection we provide a systematic comparison with previous self-supervised methods for node representation learning, including DGI [48], MVGRL [15], GRACE [57], GCA [58] and BGRL [39], and highlight the merits of CCA-SSG.”
• SimGRACE: A Simple Framework for Graph Contrastive Learning without Data Augmentation (2022, background): “(3) SimGRACE focuses on graph-level representation learning while BGRL and MERIT only work in node-level tasks.”
• OGB-LSC: A Large-Scale Challenge for Machine Learning on Graphs (2021, methodology): “, 2017) but trained it with self-supervised contrastive learning on unlabeled paper nodes (Thakoor et al., 2021).”
• Augmentation-Free Self-Supervised Learning on Graphs (2021, methodology): “More precisely, these encoders 1 AFGRL adopts the architecture of BGRL (Thakoor et al. 2021), which is slightly modified from BYOL (Grill et al. 2020).”
• Diffusion Models: A Comprehensive Survey of Methods and Applications (2022, background): “Graph Neural Networks [85, 251, 266, 285] and corresponding representation learning [86] techniques have achieved great success [14, 231, 250, 258, 264, 288] in many areas, including modeling molecule graph in various tasks ranging from property prediction [59, 71] to molecule generation [105, 111, ”