The field of graph neural networks (GNNs) is rapidly advancing, with significant developments in recent weeks. Researchers are exploring new architectures, such as the Graph Tsetlin Machine, which combines the strengths of graph neural networks and Tsetlin machines to achieve state-of-the-art results on benchmark datasets. Other works focus on improving the robustness of GNNs to noisy or missing data, such as the proposal of GraphALP, a graph augmentation framework that leverages large language models and pseudo-labeling to alleviate class imbalance and label noise. Additionally, there is a growing interest in applying GNNs to real-world problems, including recommendation systems, where researchers are investigating the use of graph convolutions in the testing phase to improve efficiency and scalability. Noteworthy papers include the proposal of ReDiSC, a reparameterized masked diffusion model for structured node classification, and GLANCE, a graph logic attention network with cluster enhancement for heterophilous graph representation learning. These advances have the potential to significantly impact various applications, from social network analysis to decision-making in high-stakes domains.
Advances in Graph Neural Networks and Related Techniques
Sources
Scalable Attribute-Missing Graph Clustering via Neighborhood Differentiatio
Enhancing Breast Cancer Detection with Vision Transformers and Graph Neural Networks
Improving KAN with CDF normalization to quantiles
Tri-Learn Graph Fusion Network for Attributed Graph Clustering
SamGoG: A Sampling-Based Graph-of-Graphs Framework for Imbalanced Graph Classification
Dual-Center Graph Clustering with Neighbor Distribution
Robust Anomaly Detection with Graph Neural Networks using Controllability
Structural Connectome Harmonization Using Deep Learning: The Strength of Graph Neural Networks
Kolmogorov Arnold Networks (KANs) for Imbalanced Data -- An Empirical Perspective
ReDiSC: A Reparameterized Masked Diffusion Model for Scalable Node Classification with Structured Predictions
LPS-GNN : Deploying Graph Neural Networks on Graphs with 100-Billion Edges
Revisiting Graph Contrastive Learning on Anomaly Detection: A Structural Imbalance Perspective
The Tsetlin Machine Goes Deep: Logical Learning and Reasoning With Graphs
Feature Construction Using Network Control Theory and Rank Encoding for Graph Machine Learning
Disentangling Homophily and Heterophily in Multimodal Graph Clustering
1.64-Approximation for Chromatic Correlation Clustering via Chromatic Cluster LP
Graph Attention Specialized Expert Fusion Model for Node Classification: Based on Cora and Pubmed Datasets
Quantifying Holistic Review: A Multi-Modal Approach to College Admissions Prediction
Querying Graph-Relational Data
Leveraging Personalized PageRank and Higher-Order Topological Structures for Heterophily Mitigation in Graph Neural Networks
Symbolic Graph Intelligence: Hypervector Message Passing for Learning Graph-Level Patterns with Tsetlin Machines
PyG 2.0: Scalable Learning on Real World Graphs
Probabilistic Graphical Models: A Concise Tutorial
BGM-HAN: A Hierarchical Attention Network for Accurate and Fair Decision Assessment on Semi-Structured Profiles
Towards Effective Open-set Graph Class-incremental Learning
Explainable Graph Neural Networks via Structural Externalities
Logical Characterizations of GNNs with Mean Aggregation
When Noisy Labels Meet Class Imbalance on Graphs: A Graph Augmentation Method with LLM and Pseudo Label
The Best is Yet to Come: Graph Convolution in the Testing Phase for Multimodal Recommendation
GLANCE: Graph Logic Attention Network with Cluster Enhancement for Heterophilous Graph Representation Learning