Simplifying Graph Convolutional Networks

2019/02/19 by Felix Wu, Tianyi Zhang, Wu, Felix +10 · 246 citations
Computer Science · Mathematics · Physics and Astronomy · #Advanced Graph Neural Networks #Complex Network Analysis Techniques #FOS: Computer and information sciences #Machine Learning (cs.LG) #Machine Learning (stat.ML) #Topic Modeling #cs.LG #stat.ML

paper · pdf · doi:10.48550/arxiv.1902.07153

In ICML 2019. Code available at https://github.com/Tiiiger/SGC

openalex publication_date 2019/02/19 · arxiv created 2019/06/20 · arxiv updated 2019/06/21 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

Graph Convolutional Networks (GCNs) and their variants have experienced significant attention and have become the de facto methods for learning graph representations. GCNs derive inspiration primarily from recent deep learning approaches, and as a result, may inherit unnecessary complexity and redundant computation. In this paper, we reduce this excess complexity through successively removing nonlinearities and collapsing weight matrices between consecutive layers. We theoretically analyze the resulting linear model and show that it corresponds to a fixed low-pass filter followed by a linear classifier. Notably, our experimental evaluation demonstrates that these simplifications do not negatively impact accuracy in many downstream applications. Moreover, the resulting model scales to larger datasets, is naturally interpretable, and yields up to two orders of magnitude speedup over FastGCN.

Citations

• The perceptron: A probabilistic model for information storage and organization in the brain.
• On the limited memory BFGS method for large scale optimization
• Long Short-Term Memory
• Backpropagation Applied to Handwritten Zip Code Recognition
• Inductive Representation Learning on Large Graphs
• Graph Neural Networks: A Review of Methods and Applications
• LanczosNet: Multi-Scale Deep Graph Convolutional Networks
• FastGCN: Fast Learning with Graph Convolutional Networks via Importance Sampling
• Detecting Functionality-Specific Vulnerabilities via Retrieving Individual Functionality-Equivalent APIs in Open-Source Repositories
• A Comprehensive Survey on Graph Neural Networks

Cited by

• Interpretable and Adaptive Node Classification on Heterophilic Graphs via Combinatorial Scoring and Hybrid Learning
• PEANUT: Perturbations by Eigenvector Alignment for Attacking Graph Neural Networks Under Topology-Driven Message Passing
• A Novel Graph-Sequence Learning Model for Inductive Text Classification
• Jensen-Shannon Divergence Message-Passing for Rich-Text Graph Representation Learning
• APC-GNN++: An Adaptive Patient-Centric GNN with Context-Aware Attention and Mini-Graph Explainability for Diabetes Classification
• Feature-Centric Unsupervised Node Representation Learning Without Homophily Assumption
• ATLAS: Adaptive Topology-based Learning at Scale for Homophilic and Heterophilic Graphs
• ParaFormer: A Generalized PageRank Graph Transformer for Graph Representation Learning
• Citation importance-aware document representation learning for large-scale science mapping
• View Space: Learning Representation across Arbitrary Graphs
• Can TabPFN Compete with GNNs for Node Classification via Graph Tabularization?
• Solving Oversmoothing in GNNs via Nonlocal Message Passing: Algebraic Smoothing and Depth Scalability
• PR-CapsNet: Pseudo-Riemannian Capsule Network with Adaptive Curvature Routing for Graph Learning
• Towards Robust DeepFake Detection under Unstable Face Sequences: Adaptive Sparse Graph Embedding with Order-Free Representation and Explicit Laplacian Spectral Prior
• Measuring Over-smoothing beyond Dirichlet energy
• Learning the Cosmic Web: Graph-based Classification of Simulated Galaxies by their Dark Matter Environments
• Conformal Correction for Efficiency May be at Odds with Entropy
• Structured Spectral Reasoning for Frequency-Adaptive Multimodal Recommendation
• Polynomial Neural Sheaf Diffusion: A Spectral Filtering Approach on Cellular Sheaves
• A3T-GCN for FTSE100 Components Price Forecasting
• E2E-GRec: An End-to-End Joint Training Framework for Graph Neural Networks and Recommender Systems
• Cross-Contrastive Clustering for Multimodal Attributed Graphs with Dual Graph Filtering
• Rethinking Semi-Supervised Node Classification with Self-Supervised Graph Clustering
• Rethinking Message Passing Neural Networks with Diffusion Distance-guided Stress Majorization
• Classification EM-PCA for clustering and embedding
• Gauge-Equivariant Graph Networks via Self-Interference Cancellation
• Connectivity-Guided Sparsification of 2-FWL GNNs: Preserving Full Expressivity with Improved Efficiency
• Hierarchical Frequency-Decomposition Graph Neural Networks for Road Network Representation Learning
• VISAGNN: Versatile Staleness-Aware Efficient Training on Large-Scale Graphs
• GraphToxin: Reconstructing Full Unlearned Graphs from Graph Unlearning
• Beyond Fixed Depth: Adaptive Graph Neural Networks for Node Classification Under Varying Homophily
• Adaptive Initial Residual Connections for GNNs with Theoretical Guarantees
• How Wide and How Deep? Mitigating Over-Squashing of GNNs via Channel Capacity Constrained Estimation
• Fixed-point graph convolutional networks against adversarial attacks
• Do We Really Need Adaptive Global Spatial Attention for Traffic Forecasting?
• Parameter-Free Hypergraph Neural Network for Few-Shot Node Classification
• Adaptive Graph Mixture of Residual Experts: Unsupervised Learning on Diverse Graphs with Heterogeneous Specialization
• Towards Unsupervised Open-Set Graph Domain Adaptation via Dual Reprogramming
• Fighter: Unveiling the Graph Convolutional Nature of Transformers in Time Series Modeling
• Rethinking Toxicity Evaluation in Large Language Models: A Multi-Label Perspective
• RoBCtrl: Attacking GNN-Based Social Bot Detectors via Reinforced Manipulation of Bots Control Interaction
• Graph Few-Shot Learning via Adaptive Spectrum Experts and Cross-Set Distribution Calibration
• LLMAtKGE: Large Language Models as Explainable Attackers against Knowledge Graph Embeddings
• Lighter-X: An Efficient and Plug-and-play Strategy for Graph-based Recommendation through Decoupled Propagation
• Learning Joint Embeddings of Function and Process Call Graphs for Malware Detection
• Analytical Survey of Learning with Low-Resource Data: From Analysis to Investigation
• Analyzing the Effect of Embedding Norms and Singular Values to Oversmoothing in Graph Neural Networks
• GILT: An LLM-Free, Tuning-Free Graph Foundational Model for In-Context Learning
• Understanding Collective Social Behavior in OSS Communities: A Co-editing Network Analysis of Activity Cascades
• Pure Node Selection for Imbalanced Graph Node Classification
• Node Classification via Simplicial Interaction with Augmented Maximal Clique Selection
• Quantum Graph Attention Network: A Novel Quantum Multi-Head Attention Mechanism for Graph Learning
• Attention Beyond Neighborhoods: Reviving Transformer for Graph Clustering
• Exploring the Global-to-Local Attention Scheme in Graph Transformers: An Empirical Study
• Towards Pre-trained Graph Condensation via Optimal Transport
• Understanding and Tackling Over-Dilution in Graph Neural Networks
• Graph Homophily Booster: Rethinking the Role of Discrete Features on Heterophilic Graphs
• Learning from Heterophilic Graphs: A Spectral Theory Perspective on the Impact of Self-Loops and Parallel Edges
• On the Impact of Downstream Tasks on Sampling and Reconstructing Noisy Graph Signals
• Why does your graph neural network fail on some graphs? Insights from exact generalisation error
• Approximate Graph Propagation Revisited: Dynamic Parameterized Queries, Tighter Bounds and Dynamic Updates
• Two Facets of the Same Optimization Coin: Model Degradation and Representation Collapse in Graph Foundation Models
• Asynchronous Message Passing for Addressing Oversquashing in Graph Neural Networks
• Graph Unlearning: Efficient Node Removal in Graph Neural Networks
• Task-Aware Adaptive Modulation: A Replay-Free and Resource-Efficient Approach For Continual Graph Learning
• GegenNet: Spectral Convolutional Neural Networks for Link Sign Prediction in Signed Bipartite Graphs
• Parameter-Free Structural-Diversity Message Passing for Graph Neural Networks
• Intellectual Property in Graph-Based Machine Learning as a Service: Attacks and Defenses
• IsoSched: Preemptive Tile Cascaded Scheduling of Multi-DNN via Subgraph Isomorphism
• Memorization in Graph Neural Networks
• Semi-Supervised Node Classification on Graphs: Markov Random Fields vs. Graph Neural Networks
• GraphFL: A Federated Learning Framework for Semi-Supervised Node Classification on Graphs
• On the Interplay between Graph Structure and Learning Algorithms in Graph Neural Networks
• Multi-view Graph Condensation via Tensor Decomposition
• A Graph Convolutional Network Composition Framework for Semi-supervised Classification
• Adversarial Graph Disentanglement
• FreeGAD: A Training-Free yet Effective Approach for Graph Anomaly Detection
• Beyond Semantic Understanding: Preserving Collaborative Frequency Components in LLM-based Recommendation
• Decoupling feature propagation from the design of graph auto-encoders
• Human Action Recognition with Multi-Laplacian Graph Convolutional Networks
• Fast Haar Transforms for Graph Neural Networks
• Over-Squashing in GNNs and Causal Inference of Rewiring Strategies
• Graph Neural Networks Inspired by Classical Iterative Algorithms
• BlindGuard: Safeguarding LLM-based Multi-Agent Systems under Unknown Attacks
• Geometry-Aware Spiking Graph Neural Network
• Transferring Social Network Knowledge from Multiple GNN Teachers to Kolmogorov-Arnold Networks
• Expanding Label Sets for Graph Convolutional Networks
• A Targeted Universal Attack on Graph Convolutional Network
• Adaptive Heterogeneous Graph Neural Networks: Bridging Heterophily and Heterogeneity
• When Deepfake Detection Meets Graph Neural Network:a Unified and Lightweight Learning Framework
• Fisher-Bures Adversary Graph Convolutional Networks
• Complete the Missing Half: Augmenting Aggregation Filtering with Diversification for Graph Convolutional Networks
• Leveraging Personalized PageRank and Higher-Order Topological Structures for Heterophily Mitigation in Graph Neural Networks
• Online Continual Graph Learning
• Federated Graph Unlearning
• MPFSR-Enhanced GNNs: Spectral Graph Neural Networks Enhancement Through Learnable Multiple-Parameter Graph Fractional Fourier Transforms
• Simple Truncated SVD based Model for Node Classification on Heterophilic Graphs
• Spectral Geometric Matrix Completion
• GraphTorque: Torque-Driven Rewiring Graph Neural Network
• Fusion with Hierarchical Graphs for Mulitmodal Emotion Recognition
• Active Learning for Graph Neural Networks via Node Feature Propagation
• Encoder-Inverter Framework for Seismic Acoustic Impedance Inversion
• DyGCN: Dynamic Graph Embedding with Graph Convolutional Network
• Non-Recursive Graph Convolutional Networks
• LookHops: light multi-order convolution and pooling for graph classification
• A Hierarchy of Graph Neural Networks Based on Learnable Local Features
• ProGMLP: A Progressive Framework for GNN-to-MLP Knowledge Distillation with Efficient Trade-offs
• Redundancy-Free Computation Graphs for Graph Neural Networks
• FedSA-GCL: A Semi-Asynchronous Federated Graph Learning Framework with Personalized Aggregation and Cluster-Aware Broadcasting
• Scaling Up Graph Neural Networks Via Graph Coarsening
• Research on the application of graph data structure and graph neural network in node classification/clustering tasks
• Graph Neighborhood Attentive Pooling
• Graph-Preserving Grid Layout: A Simple Graph Drawing Method for Graph Classification using CNNs
• Jointly Learnable Data Augmentations for Self-Supervised GNNs
• LPS-GNN : Deploying Graph Neural Networks on Graphs with 100-Billion Edges
• EEG-Based Emotion Recognition Using Regularized Graph Neural Networks
• Unsupervised Deep Manifold Attributed Graph Embedding
• Tracing the Propagation Path: A Flow Perspective of Representation Learning on Graphs
• Robust Anomaly Detection with Graph Neural Networks using Controllability
• Graph Neural Networks with Feature and Structure Aware Random Walk
• When Product Search Meets Collaborative Filtering: A Hierarchical Heterogeneous Graph Neural Network Approach
• GCN for HIN via Implicit Utilization of Attention and Meta-paths
• Meta-GNN: On Few-shot Node Classification in Graph Meta-learning
• Graph-Revised Convolutional Network
• Improving Graph Neural Networks with Simple Architecture Design
• A Survey on Recommender Systems Using Graph Neural Network
• A Survey on Text Classification: From Shallow to Deep Learning
• Simplification of Graph Convolutional Networks: A Matrix Factorization-based Perspective
• Deconvolutional Networks on Graph Data
• Graph World Model
• Graph Star Net for Generalized Multi-Task Learning
• Adapting to Heterophilic Graph Data with Structure-Guided Neighbor Discovery
• NLGCL: Naturally Existing Neighbor Layers Graph Contrastive Learning for Recommendation
• Scalable Attribute-Missing Graph Clustering via Neighborhood Differentiation
• Learnable quantum spectral filters for hybrid graph neural networks
• Trainable Class Prototypes for Few-Shot Learning
• Simplifying Graph Kernels for Efficient
• Graph-MLP: Node Classification without Message Passing in Graph
• Fully Linear Graph Convolutional Networks for Semi-Supervised Learning and Clustering
• SoGCN: Second-Order Graph Convolutional Networks
• Co-embedding of Nodes and Edges with Graph Neural Networks
• PINE: Universal Deep Embedding for Graph Nodes via Partial Permutation Invariant Set Functions
• Single-Layer Graph Convolutional Networks For Recommendation
• Convergent Privacy Framework for Multi-layer GNNs through Contractive Message Passing
• CoATA: Effective Co-Augmentation of Topology and Attribute for Graph Neural Networks
• Enhancing Homophily-Heterophily Separation: Relation-Aware Learning in Heterogeneous Graphs
• PlaceFM: A Training-free Geospatial Foundation Model of Places using Large-Scale Point of Interest Data
• Higher-Order Graph Databases
• SaGIF: Improving Individual Fairness in Graph Neural Networks via Similarity Encoding
• Learning Parametrised Graph Shift Operators
• Revisiting Graph Convolutional Network on Semi-Supervised Node Classification from an Optimization Perspective
• Towards a deeper GCN: Alleviate over-smoothing with iterative training and fine-tuning
• Adversarial Examples on Graph Data: Deep Insights into Attack and Defense
• On the Equivalence of Decoupled Graph Convolution Network and Label Propagation
• Totally Deep Support Vector Machines
• Label-GCN: An Effective Method for Adding Label Propagation to Graph Convolutional Networks
• Return of ChebNet: Understanding and Improving an Overlooked GNN on Long Range Tasks
• Semi-supervised Graph Anomaly Detection via Robust Homophily Learning
• Beltrami Flow and Neural Diffusion on Graphs
• Equivariance Everywhere All At Once: A Recipe for Graph Foundation Models
• Uncertainty-Aware Graph Neural Networks: A Multi-Hop Evidence Fusion Approach
• Taxonomy of reduction matrices for Graph Coarsening
• LightKG: Efficient Knowledge-Aware Recommendations with Simplified GNN Architecture
• Resilient UAV Swarm Communications with Graph Convolutional Neural Network
• Topology-aware Neural Flux Prediction Guided by Physics
• iN2V: Bringing Transductive Node Embeddings to Inductive Graphs
• The Oversmoothing Fallacy: A Misguided Narrative in GNN Research
• Fundamental Limits of Deep Graph Convolutional Networks
• Are Powerful Graph Neural Nets Necessary? A Dissection on Graph Classification
• Deep Graph Library: A Graph-Centric, Highly-Performant Package for Graph Neural Networks
• Interpreting and Unifying Graph Neural Networks with An Optimization Framework
• L²-GCN: Layer-Wise and Learned Efficient Training of Graph Convolutional Networks
• Deep graph learning for semi-supervised classification
• Reconstructing nodal pressures in water distribution systems with graph neural networks
• Deep Learning on Graphs: A Survey
• Learning Binarized Representations with Pseudo-positive Sample Enhancement for Efficient Graph Collaborative Filtering
• Unlearning Inversion Attacks for Graph Neural Networks
• On the Stability of Graph Convolutional Neural Networks: A Probabilistic Perspective
• Hyperbolic-PDE GNN: Spectral Graph Neural Networks in the Perspective of A System of Hyperbolic Partial Differential Equations
• A Benchmark Dataset for Graph Regression with Homogeneous and Multi-Relational Variants
• From Aggregation to Selection: User-Validated Distributed Social Recommendation
• Aggregation Buffer: Revisiting DropEdge with a New Parameter Block
• Cooperation of Experts: Fusing Heterogeneous Information with Large Margin
• Graph Wave Networks
• Towards Effective Federated Multimodal Graph Learning via Navigating Multifaceted Heterogeneity
• Deep Learning Based Text Classification: A Comprehensive Review
• Lightweight Embeddings with Graph Rewiring for Collaborative Filtering
• Learn Beneficial Noise as Graph Augmentation
• Privacy-Preserving Representation Learning on Graphs: A Mutual Information Perspective
• CoRe-GNN: Multilevel Message passing on Coarsened graphs
• Empowering Credit Risk Detection in Weixin Pay with Billion-Scale Deep Graph Learning
• DAM-GT: Dual Positional Encoding-Based Attention Masking Graph Transformer for Node Classification
• InfoGCL: Information-Aware Graph Contrastive Learning
• Only Large Weights (And Not Skip Connections) Can Prevent the Perils of Rank Collapse
• A Point Cloud Generative Model via Tree-Structured Graph Convolutions for 3D Brain Shape Reconstruction
• p-Laplacian Based Graph Neural Networks
• Partition-wise Graph Filtering: A Unified Perspective Through the Lens of Graph Coarsening
• TxPert: Leveraging Biochemical Relationships for Out-of-Distribution Transcriptomic Perturbation Prediction
• Nonparametric Teaching for Graph Property Learners
• Unlearning Algorithmic Biases over Graphs
• A Hard Label Black-box Adversarial Attack Against Graph Neural Networks
• Towards Deeper Graph Neural Networks with Differentiable Group Normalization
• Adaptive Tokenization: On the Hop-Overpriority Problem in Tokenized Graph Learning Models
• AH-UGC: Adaptive and Heterogeneous-Universal Graph Coarsening
• Fast RoPE Attention: Combining the Polynomial Method and Fast Fourier Transform
• Evaluating Deep Graph Neural Networks
• Graph Neural Networks Exponentially Lose Expressive Power for Node Classification
• Contrastive and Generative Graph Convolutional Networks for Graph-based Semi-Supervised Learning
• QSeer: A Quantum-Inspired Graph Neural Network for Parameter Initialization in Quantum Approximate Optimization Algorithm Circuits
• Fair Representation Learning for Continuous Sensitive Attributes using Expectation of Integral Probability Metrics
• Wide & Deep Learning for Node Classification
• On the Stability of Graph Convolutional Neural Networks under Edge Rewiring
• Isometric Transformation Invariant and Equivariant Graph Convolutional Networks
• SAC: Accelerating and Structuring Self-Attention via Sparse Adaptive Connection
• Toward Data-centric Directed Graph Learning: An Entropy-driven Approach
• Graph Spectral Filtering with Chebyshev Interpolation for Recommendation
• ExpandNets: Linear Over-parameterization to Train Compact Convolutional Networks
• Self-Supervised Hypergraph Convolutional Networks for Session-based Recommendation
• Unified Multi-Domain Graph Pre-training for Homogeneous and Heterogeneous Graphs via Domain-Specific Expert Encoding
• Understanding GNNs and Homophily in Dynamic Node Classification
• Quantifying the Noise of Structural Perturbations on Graph Adversarial Attacks
• A Generative Graph Contrastive Learning Model with Global Signal
• Modeling multi-scale data via a network of networks
• Neural Bandit with Arm Group Graph
• Mask-GVAE: Blind Denoising Graphs via Partition
• Continuous-Depth Neural Models for Dynamic Graph Prediction
• Adversarial Attack Framework on Graph Embedding Models with Limited Knowledge
• ScaleGNN: Towards Scalable Graph Neural Networks via Adaptive High-order Neighboring Feature Fusion
• PushNet: Efficient and Adaptive Neural Message Passing
• Anomaly Detection on Attributed Networks via Contrastive Self-Supervised Learning
• GraphEdge: Dynamic Graph Partition and Task Scheduling for GNNs Computing in Edge Network
• Heterogeneous networks in drug-target interaction prediction
• Scalable Graph Neural Networks for Heterogeneous Graphs
• Fairness-Aware Node Representation Learning
• A Comparison of Neural Network Architectures for Data-Driven Reduced-Order Modeling
• GraphAIR: Graph representation learning with neighborhood aggregation and interaction
• Simplifying Graph Convolutional Networks with Redundancy-Free Neighbors
• The Power of Graph Convolutional Networks to Distinguish Random Graph\n Models: Short Version
• Defect Prediction With Semantics and Context Features of Codes Based on Graph Representation Learning
• Parallel Multi-Graph Convolution Network For Metro Passenger Volume Prediction
• Graph Learning at Scale: Characterizing and Optimizing Pre-Propagation GNNs
• Federated Spectral Graph Transformers Meet Neural Ordinary Differential Equations for Non-IID Graphs
• Extended Short- and Long-Range Mesh Learning for Fast and Generalized Garment Simulation
• Progress in \cal CP violating top-Higgs coupling at the LHC with Machine Learning
• H³GNNs: Harmonizing Heterophily and Homophily in GNNs via Joint Structural Node Encoding and Self-Supervised Learning
• Towards Unbiased Federated Graph Learning: Label and Topology Perspectives
• Federated Prototype Graph Learning
• SAFT: Structure-aware Transformers for Textual Interaction Classification

Related

• Simple Graph Convolutional Networks
• Simple and Deep Graph Convolutional Networks
• On Provable Benefits of Depth in Training Graph Convolutional Networks
• Simple yet Effective Gradient-Free Graph Convolutional Networks
• L²-GCN: Layer-Wise and Learned Efficient Training of Graph Convolutional Networks