LanczosNet: Multi-Scale Deep Graph Convolutional Networks

We propose the Lanczos network (LanczosNet), which uses the Lanczos algorithm to construct low rank approximations of the graph Laplacian for graph convolution. Relying on the tridiagonal decomposition of the Lanczos algorithm, we not only efficiently exploit multi-scale information via fast approximated computation of matrix power but also design learnable spectral filters. Being fully differentiable, LanczosNet facilitates both graph kernel learning as well as learning node embeddings. We show the connection between our LanczosNet and graph based manifold learning methods, especially the diffusion maps. We benchmark our model against several recent deep graph networks on citation networks and QM8 quantum chemistry dataset. Experimental results show that our model achieves the state-of-the-art performance in most tasks. Code is released at: \url{this https URL}.

• Publication year

  2019

• Venue

  International Conference on Learning Representations

• Publication date

  2019-01-06

• Fields of study

  Mathematics, Computer Science

• Identifiers
  arXiv 1901.01484
• External record

  Open on Semantic Scholar

• Source metadata

  Semantic Scholar

• No claims are published for this paper.

• No concepts are published for this paper.

• Spectral graph theory

  2019cited by this paper
• Relational inductive biases, deep learning, and graph networks

  2018cited by this paper
• Multi-scale kernels for Nyström based extension schemes

  2018cited by this paper
• FastGCN: Fast Learning with Graph Convolutional Networks via Importance Sampling

  2018cited by this paper
• Graph networks as learnable physics engines for inference and control

  2018cited by this paper
• Graph Partition Neural Networks for Semi-Supervised Classification

  2018cited by this paper
• Inference in Probabilistic Graphical Models by Graph Neural Networks

  2018cited by this paper
• Junction Tree Variational Autoencoder for Molecular Graph Generation

  2018cited by this paper
• Learning Deep Generative Models of Graphs

  2018cited by this paper
• Iteration Method for the Solution of the Eigenvalue Problem of Linear Differential and Integral Operators

  2018cited by this paper
• NerveNet: Learning Structured Policy with Graph Neural Networks

  2018cited by this paper
• Deep Parametric Continuous Convolutional Neural Networks

  2018cited by this paper
• 3D Graph Neural Networks for RGBD Semantic Segmentation

  2017cited by this paper
• Neural Message Passing for Quantum Chemistry

  2017influential reference
• Situation Recognition with Graph Neural Networks

  2017cited by this paper
• Learning Graph Representations with Embedding Propagation

  2017cited by this paper
• Inductive Representation Learning on Large Graphs

  2017cited by this paper
• MoleculeNet: a benchmark for molecular machine learning

  2017cited by this paper
• Graph Signal Processing: Overview, Challenges, and Applications

  2017cited by this paper
• Automatic differentiation in PyTorch

  2017cited by this paper
• Graph Attention Networks

  2017influential reference
• Community Detection with Graph Neural Networks

  2017cited by this paper
• Few-Shot Learning with Graph Neural Networks

  2017cited by this paper
• Stochastic Training of Graph Convolutional Networks with Variance Reduction

  2017cited by this paper
• Semi-Supervised Classification with Graph Convolutional Networks

  2016influential reference
• Geometric Deep Learning: Going beyond Euclidean data

  2016cited by this paper
• Revisiting Semi-Supervised Learning with Graph Embeddings

  2016cited by this paper
• Learning Convolutional Neural Networks for Graphs

  2016cited by this paper
• node2vec: Scalable Feature Learning for Networks

  2016cited by this paper
• Convolutional Neural Networks on Graphs with Fast Localized Spectral Filtering

  2016influential reference
• Discriminative Embeddings of Latent Variable Models for Structured Data

  2016cited by this paper
• Accelerated filtering on graphs using Lanczos method

  2015cited by this paper
• Convolutional Networks on Graphs for Learning Molecular Fingerprints

  2015cited by this paper
• Gated Graph Sequence Neural Networks

  2015influential reference
• Deep Convolutional Networks on Graph-Structured Data

  2015cited by this paper
• Diffusion-Convolutional Neural Networks

  2015influential reference
• Electronic spectra from TDDFT and machine learning in chemical space.

  2015influential reference
• Dynamics-Adapted Cone Kernels

  2014cited by this paper
• Analog forecasting with dynamics-adapted kernels

  2014cited by this paper
• Empirical intrinsic geometry for nonlinear modeling and time series filtering

  2013cited by this paper
• Spectral Networks and Locally Connected Networks on Graphs

  2013cited by this paper
• The emerging field of signal processing on graphs: Extending high-dimensional data analysis to networks and other irregular domains

  2012cited by this paper
• The conference paper

  2011cited by this paper
• Detecting intrinsic slow variables in stochastic dynamical systems by anisotropic diffusion maps

  2009cited by this paper
• Wavelets on Graphs via Spectral Graph Theory

  2009cited by this paper
• The Graph Neural Network Model

  2009cited by this paper
• Diffusion Interpretation of Nonlocal Neighborhood Filters for Signal Denoising

  2009cited by this paper
• Non-linear independent component analysis with diffusion maps

  2008cited by this paper
• Graph Kernels

  2008cited by this paper
• Metric Learning for Kernel Regression

  2007cited by this paper
• A tutorial on spectral clustering

  2007cited by this paper
• Numerical Recipes 3rd Edition: The Art of Scientific Computing

  2007cited by this paper
• Diffusion maps

  2006influential reference
• Manifold Regularization: A Geometric Framework for Learning from Labeled and Unlabeled Examples

  2006cited by this paper
• --1 CONTENTS

  2006cited by this paper
• From graph to manifold Laplacian: The convergence rate

  2006cited by this paper
• Diffusion Maps, Spectral Clustering and Eigenfunctions of Fokker-Planck Operators

  2005cited by this paper
• Diffusion Kernels on Statistical Manifolds

  2005cited by this paper
• Geometric diffusions as a tool for harmonic analysis and structure definition of data: multiscale methods.

  2005cited by this paper
• Self-Tuning Spectral Clustering

  2004cited by this paper
• Hessian eigenmaps: Locally linear embedding techniques for high-dimensional data

  2003cited by this paper
• Laplacian Eigenmaps and Spectral Techniques for Embedding and Clustering

  2001cited by this paper
• A global geometric framework for nonlinear dimensionality reduction.

  2000cited by this paper
• Methods of information geometry

  2000cited by this paper
• Nonlinear dimensionality reduction by locally linear embedding.

  2000cited by this paper
• Low-Rank Matrix Approximation Using the Lanczos Bidiagonalization Process with Applications

  1999cited by this paper
• Recursive Distributed Representations

  1990cited by this paper
• Scale-Space Filtering

  1983cited by this paper
• The Symmetric Eigenvalue Problem

  1981cited by this paper
• A Global Geometric Framework for Nonlinear Dimensionality Reduction

  year unknowncited by this paper

• BPE-MTSNet: Balanced Performance and Efficiency in Multivariate Spatiotemporal Time Series Forecasting

  2026cites this paper
• Enhancing deep hashing with graph filters and autoencoder-based embeddings

  2026cites this paper
• Stable Representation Learning via Generalized Learnable Graph Scattering Transform

  2026cites this paper
• Let Invariant Learning Inspire Neighbor-Shift Generalization on Graphs

  2026cites this paper
• A federated graph learning method to realize multi-party collaboration for molecular discovery

  2026influential citation
• A Graph Laplacian Eigenvector-based Pre-training Method for Graph Neural Networks

  2025cites this paper
• FaLa: feature-augmented location-aware transformer network for graph representation learning

  2025cites this paper
• Beyond Message Passing: Neural Graph Pattern Machine

  2025cites this paper
• Informed machine learning for complex data

  2025cites this paper
• Topological Fusion Model for Molecular Property Prediction

  2025cites this paper
• JFRFFNet: A Data–Model Co-Driven Graph Signal Denoising Model With Partial Prior Information

  2025cites this paper
• Dynamic Multiple-Parameter Joint Time-Vertex Fractional Fourier Transform and its Intelligent Filtering Methods

  2025cites this paper
• Long-Range Graph Wavelet Networks

  2025cites this paper
• Piecewise Constant Spectral Graph Neural Network

  2025cites this paper
• JointSTNet: Joint Pre-Training for Spatial-Temporal Traffic Forecasting

  2025cites this paper
• Polynomial Neural Sheaf Diffusion: A Spectral Filtering Approach on Cellular Sheaves

  2025cites this paper
• Survey on Graph Neural Networks

  2024cites this paper
• Graph Denoising With Framelet Regularizers

  2024cites this paper
• Human action recognition using ST-GCNs for blind accessible theatre performances

  2024cites this paper
• Residual Attention Augmentation Graph Neural Network for Improved Node Classification

  2024cites this paper
• Rethinking Spectral Graph Neural Networks with Spatially Adaptive Filtering

  2024cites this paper
• Point Cloud Denoising With Fine-Granularity Dynamic Graph Convolutional Networks

  2024cites this paper
• Research on Multi-Scale Spatio-Temporal Graph Convolutional Human Behavior Recognition Method Incorporating Multi-Granularity Features

  2024cites this paper
• Strong Transitivity Relations and Graph Neural Networks

  2024cites this paper
• Graph Edge Representation via Tensor Product Graph Convolutional Representation

  2024influential citation
• Spatio-spectral graph neural operator for solving computational mechanics problems on irregular domain and unstructured grid

  2024cites this paper
• CKGConv: General Graph Convolution with Continuous Kernels

  2024cites this paper
• Machine learning applications for thermochemical and kinetic property prediction

  2024cites this paper
• Sub-sampling graph neural networks for genomic prediction of quantitative phenotypes

  2024cites this paper
• Structure-Aware Consensus Network on Graphs with Few Labeled Nodes

  2024influential citation
• Graph Involutional Networks with Dynamic Feature Fusion for Skeleton-Based Action Recognition

  2024cites this paper
• Spatio-Spectral Graph Neural Networks

  2024cites this paper
• PhyMPGN: Physics-encoded Message Passing Graph Network for spatiotemporal PDE systems

  2024cites this paper
• Cross-Modal Learning for Chemistry Property Prediction: Large Language Models Meet Graph Machine Learning

  2024cites this paper
• Graph Adaptive Attention Network with Cross-Entropy

  2024cites this paper
• Graph Linear Canonical Transform: Definition, Vertex-Frequency Analysis and Filter Design

  2024influential citation
• Graph Neural Aggregation-diffusion with Metastability

  2024cites this paper
• Non-convolutional Graph Neural Networks

  2024cites this paper
• Benchmarking Spectral Graph Neural Networks: A Comprehensive Study on Effectiveness and Efficiency

  2024cites this paper
• Manifold GCN: Diffusion-based Convolutional Neural Network for Manifold-valued Graphs

  2024cites this paper
• Path Integral Based Convolution and Pooling for Heterogeneous Graph Neural Networks

  2023cites this paper
• Molecular Property Prediction of Modified Gedunin Using Machine Learning

  2023cites this paper
• Specformer: Spectral Graph Neural Networks Meet Transformers

  2023cites this paper
• Hyperbolic Temporal Network Embedding

  2023cites this paper
• Fully Linear Graph Convolutional Networks for Semi-Supervised and Unsupervised Classification

  2023cites this paper
• Graph Neural Networks With High-Order Polynomial Spectral Filters

  2023influential citation
• Bi-level Contrastive Learning for Knowledge-Enhanced Molecule Representations

  2023cites this paper
• Towards Understanding the Generalization of Graph Neural Networks

  2023cites this paper
• LD2: Scalable Heterophilous Graph Neural Network with Decoupled Embeddings

  2023influential citation
• Graph Neural Networks with Non-Recursive Message Passing

  2023cites this paper
• Multi-scale Directed Graph Convolution Neural Network for Node Classification Task

  2023cites this paper
• Feature Explanation for Robust Trajectory Prediction

  2023cites this paper
• Global disentangled graph convolutional neural network based on a graph topological metric

  2023cites this paper
• From Continuous Dynamics to Graph Neural Networks: Neural Diffusion and Beyond

  2023cites this paper
• PREM: A Simple Yet Effective Approach for Node-Level Graph Anomaly Detection

  2023cites this paper
• A Model-Agnostic Graph Neural Network for Integrating Local and Global Information

  2023cites this paper
• Graph Convolutional Network Design for Node Classification Accuracy Improvement

  2023cites this paper
• ESS MS-G3D: extension and supplement shift MS-G3D network for the assessment of severe mental retardation

  2023cites this paper
• Self-Adaptive Graph With Nonlocal Attention Network for Skeleton-Based Action Recognition

  2023cites this paper
• Empowering Simple Graph Convolutional Networks

  2023cites this paper
• A Survey on Spectral Graph Neural Networks

  2023cites this paper
• Double-head transformer neural network for molecular property prediction

  2023cites this paper
• Theater Aid System for the Visually Impaired Through Transfer Learning of Spatio-Temporal Graph Convolution Networks

  2023cites this paper
• Graph Neural Point Process for Temporal Interaction Prediction

  2023cites this paper
• Gode - Integrating Biochemical Knowledge Graph into Pre-training Molecule Graph Neural Network

  2023cites this paper
• A comprehensive review of graph convolutional networks: approaches and applications

  2023cites this paper
• Adaptive self-propagation graph convolutional network for recommendation

  2023cites this paper
• On Addressing the Limitations of Graph Neural Networks

  2023cites this paper
• Mixed-Integer Optimisation of Graph Neural Networks for Computer-Aided Molecular Design

  2023cites this paper
• NGAT: attention in breadth and depth exploration for semi-supervised graph representation learning

  2022cites this paper
• Scalable self-supervised graph representation learning via enhancing and contrasting subgraphs

  2022cites this paper
• Multi-stream adaptive 3D attention graph convolution network for skeleton-based action recognition

  2022cites this paper
• Multi-Branch Spatial-Temporal Attention Graph Convolution Network for Skeleton-based Action Recognition

  2022cites this paper
• Hierarchical Neighbors Embedding

  2022cites this paper
• How Powerful is Implicit Denoising in Graph Neural Networks

  2022cites this paper
• GPNet: Simplifying Graph Neural Networks via Multi-channel Geometric Polynomials

  2022cites this paper
• Molformer: Large Scale Chemical Language Representations Capture Molecular Structure and Properties

  2022cites this paper
• Improving Node Classification through Convolutional Networks Built on Enhanced Message-Passing Graph

  2022cites this paper
• Graph Neural Network-Based Node Deployment for Throughput Enhancement

  2022cites this paper
• From Local to Global: Spectral-Inspired Graph Neural Networks

  2022cites this paper
• Construction Safety Risk Model with Construction Accident Network: A Graph Convolutional Network Approach

  2022cites this paper
• Deep learning methods for molecular representation and property prediction.

  2022cites this paper
• Learnable Filters for Geometric Scattering Modules

  2022cites this paper
• GRAND++: Graph Neural Diffusion with A Source Term

  2022cites this paper
• Semantic-guided multi-scale human skeleton action recognition

  2022cites this paper
• Hierarchical Diffusion Scattering Graph Neural Network

  2022cites this paper
• Live Stream Temporally Embedded 3D Human Body Pose and Shape Estimation

  2022influential citation
• Graph neural networks for materials science and chemistry

  2022cites this paper
• Beat Wash-Sale Tax with Multigraph Convolutional Neural Networks Based Trading Strategy

  2022cites this paper
• Spectral Graph Convolutional Neural Networks in the Context of Regularization Theory

  2022cites this paper
• Graph Influence Network

  2022cites this paper
• Skeleton-based Action Recognition via Temporal-Channel Aggregation

  2022cites this paper
• A Simple Yet Effective Pretraining Strategy for Graph Few-shot Learning

  2022cites this paper
• Incorporating Heterophily into Graph Neural Networks for Graph Classification

  2022cites this paper
• Are Graph Convolutional Networks With Random Weights Feasible?

  2022cites this paper
• View-Consistent Heterogeneous Network on Graphs With Few Labeled Nodes

  2022cites this paper
• SIS: A new multi-scale convolutional operator

  2022cites this paper
• Overcoming Oversmoothness in Graph Convolutional Networks via Hybrid Scattering Networks

  2022cites this paper
• Graph Neural Diffusion Networks for Semi-supervised Learning

  2022cites this paper
• Multi-level disentanglement graph neural network

  2022cites this paper