Manifold GCN: Diffusion-based Convolutional Neural Network for Manifold-valued Graphs
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2024
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866908243160203264 |
|---|---|
| author | Hanik, Martin Steidl, Gabriele von Tycowicz, Christoph |
| author_facet | Hanik, Martin Steidl, Gabriele von Tycowicz, Christoph |
| contents | We propose two graph neural network layers for graphs with features in a Riemannian manifold. First, based on a manifold-valued graph diffusion equation, we construct a diffusion layer that can be applied to an arbitrary number of nodes and graph connectivity patterns. Second, we model a tangent multilayer perceptron by transferring ideas from the vector neuron framework to our general setting. Both layers are equivariant under node permutations and the feature manifold's isometries. These properties have led to a beneficial inductive bias in many deep-learning tasks. Numerical examples on synthetic data and an Alzheimer's classification application on triangle meshes of the right hippocampus demonstrate the usefulness of our new layers: While they apply to a much broader class of problems, they perform as well as or better than task-specific state-of-the-art networks. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2401_14381 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Manifold GCN: Diffusion-based Convolutional Neural Network for Manifold-valued Graphs Hanik, Martin Steidl, Gabriele von Tycowicz, Christoph Machine Learning Differential Geometry 53Z50 I.2.4 We propose two graph neural network layers for graphs with features in a Riemannian manifold. First, based on a manifold-valued graph diffusion equation, we construct a diffusion layer that can be applied to an arbitrary number of nodes and graph connectivity patterns. Second, we model a tangent multilayer perceptron by transferring ideas from the vector neuron framework to our general setting. Both layers are equivariant under node permutations and the feature manifold's isometries. These properties have led to a beneficial inductive bias in many deep-learning tasks. Numerical examples on synthetic data and an Alzheimer's classification application on triangle meshes of the right hippocampus demonstrate the usefulness of our new layers: While they apply to a much broader class of problems, they perform as well as or better than task-specific state-of-the-art networks. |
| title | Manifold GCN: Diffusion-based Convolutional Neural Network for Manifold-valued Graphs |
| topic | Machine Learning Differential Geometry 53Z50 I.2.4 |
| url | https://arxiv.org/abs/2401.14381 |