mlsanityData to 14 Oct 2026

Latent Graph Inference with Limited Supervision

arXiv 2310.04314 ↗First posted 6 Oct 2023cs.LGcited by 9

Not on the map. Why?

Abstract (arXiv)

Latent graph inference (LGI) aims to jointly learn the underlying graph structure and node representations from data features. However, existing LGI methods commonly suffer from the issue of supervision starvation, where massive edge weights are learned without semantic supervision and do not contribute to the training loss. Consequently, these supervision-starved weights, which may determine the predictions of testing samples, cannot be semantically optimal, resulting in poor generalization. In this paper, we observe that this issue is actually caused by the graph sparsification operation, which severely destroys the important connections established between pivotal nodes and labeled ones. To address this, we propose to restore the corrupted affinities and replenish the missed supervision for better LGI. The key challenge then lies in identifying the critical nodes and recovering the corrupted affinities. We begin by defining the pivotal nodes as k-hop starved nodes, which can be identified based on a given adjacency matrix. Considering the high computational burden, we further present a more efficient alternative inspired by CUR matrix decomposition. Subsequently, we eliminate the starved nodes by reconstructing the destroyed connections. Extensive experiments on representative benchmarks demonstrate that reducing the starved nodes consistently improves the performance of state-of-the-art LGI methods, especially under extremely limited supervision (6.12% improvement on Pubmed with a labeling rate of only 0.3%).

Key facts. Cited by 9 papers in the map (data through 14 Oct 2026). Builds on 31 of its 48 references that are in the map. First posted on arXiv 6 Oct 2023.

Builds on

31 of 48 references.

References · 31

  1. 2016Semi-Supervised Classification with Graph Convolutional Networkscited by 11,361
  2. 2017Graph Attention Networkscited by 8,923
  3. 2017Inductive Representation Learning on Large Graphscited by 6,320
  4. 2018Dynamic Graph CNN for Learning on Point Cloudscited by 2,223
  5. 2020Open Graph Benchmark: Datasets for Machine Learning on Graphscited by 2,163
  6. 2018Deeper Insights into Graph Convolutional Networks for Semi-Supervised Learningcited by 1,102
  7. 2021E(n) Equivariant Graph Neural Networkscited by 965
  8. 2020On the Bottleneck of Graph Neural Networks and its Practical Implicationscited by 631
  9. 2020Graph Structure Learning for Robust Graph Neural Networkscited by 371
  10. 2019A Fair Comparison of Graph Neural Networks for Graph Classificationcited by 244
  11. 2019Learning Discrete Structures for Graph Neural Networkscited by 231
  12. 2020Iterative Deep Graph Learning for Graph Neural Networks: Better and Robust Node Embeddingscited by 214
19 more
  1. 2023NodeFormer: A Scalable Graph Structure Learning Transformer for Node Classificationcited by 210
  2. 2018Adaptive Graph Convolutional Neural Networkscited by 155
  3. 2022Towards Unsupervised Deep Graph Structure Learningcited by 114
  4. 2021SLAPS: Self-Supervision Improves Structure Learning for Graph Neural Networkscited by 110
  5. 2021Learning Dynamic Graph Representation of Brain Connectome with Spatio-Temporal Attentioncited by 87
  6. 2020Differentiable Graph Module (DGM) for Graph Convolutional Networkscited by 86
  7. 2021Graph Structure Learning with Variational Information Bottleneckcited by 81
  8. 2021Multi-view Contrastive Graph Clusteringcited by 76
  9. 2019Graph-Revised Convolutional Networkcited by 59
  10. 2022Adaptive Trajectory Prediction via Transferable GNNcited by 51
  11. 2018Learning Conditioned Graph Structures for Interpretable Visual Question Answeringcited by 45
  12. 2020Latent-Graph Learning for Disease Predictioncited by 39
  13. 2014Optimal CUR Matrix Decompositionscited by 29
  14. 2021Learning to Learn Graph Topologiescited by 26
  15. 2022Latent Graph Inference using Product Manifoldscited by 23
  16. 2021Robust CUR Decomposition: Theory and Imaging Applicationscited by 20
  17. 2022Ada-NETS: Face Clustering via Adaptive Neighbour Discovery in the Structure Spacecited by 16
  18. 2022Boosting Graph Structure Learning with Dummy Nodescited by 13
  19. 2023UGSL: A Unified Framework for Benchmarking Graph Structure Learningcited by 9

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2023: 1; 2024: 2; 2025: 2; 2026: 4.

9 papers.

  1. 2026The Indra Representation Hypothesis for Multimodal Alignmentcited by 4
  2. 2026NeuroMambaLLM: Dynamic Graph Learning of fMRI Functional Connectivity in Autistic Brains Using Mamba and Language Model Reasoningcited by 3
  3. 2025Scale-Free Graph-Language Modelscited by 2
  4. 2023FALCON: Feature-Label Constrained Graph Net Collapse for Memory Efficient GNNscited by 1
  5. 2024Inference of Sequential Patterns for Neural Message Passing in Temporal Graphscited by 1
  6. 2024Boolean Product Graph Neural Networkscited by 0
  7. 2025Dependency Parsing is More Parameter-Efficient with Normalizationcited by 0
  8. 2026LLMs Underperform Graph-Based Parsers on Supervised Relation Extraction for Complex Graphscited by 0
  9. 2026A Unified Perspective for Learning Graph Representations Across Multi-Level Abstractionscited by 0