Pseudoinverse graph convolutional networks. Fast filters tailored for large eigengaps of dense graphs and hypergraphs

DOI10.1007/S10618-021-00752-WzbMath1473.68133arXiv2008.00720OpenAlexW3153142113MaRDI QIDQ2036767

Martin Stoll, Dominik Alfke

Publication date: 30 June 2021

Published in: Data Mining and Knowledge Discovery (Search for Journal in Brave)

Full work available at URL: https://arxiv.org/abs/2008.00720

zbMATH Keywords

machine learningsemi-supervised learningsignal processing on graphsgraph neural networks

Mathematics Subject Classification ID

Learning and adaptive systems in artificial intelligence (68T05) Hypergraphs (05C65) Graph theory (including graph drawing) in computer science (68R10) Graphs and linear algebra (matrices, eigenvalues, etc.) (05C50) Signal theory (characterization, reconstruction, filtering, etc.) (94A12) Density (toughness, etc.) (05C42)

Uses Software

• UCI-ml
• PyTorch
• Adam
• FastGCN

Cites Work

• Unnamed Item
• Hypergraph theory. An introduction
• Bipartite and neighborhood graphs and the spectrum of the normalized graph Laplace operator
• A Krylov--Schur Algorithm for Large Eigenproblems
• Numerical Methods for Large Eigenvalue Problems
• Nonlocal Operators with Applications to Image Processing
• Matrix Analysis
• Spectral Properties of Hypergraph Laplacian and Approximation Algorithms
• Generalizing Diffuse Interface Methods on Graphs: Nonsmooth Potentials and Hypergraphs

This page was built for publication: Pseudoinverse graph convolutional networks. Fast filters tailored for large eigengaps of dense graphs and hypergraphs