Diffusion Scattering Transforms on Graphs

Abstract

Stability is a key aspect of data analysis. In many applications, the natural notion of stability is geometric, as illustrated for example in computer vision. Scattering transforms construct deep convolutional representations which are certified stable to input deformations. This stability to deformations can be interpreted as stability with respect to changes in the metric structure of the domain. In this work, we show that scattering transforms can be generalized to non-Euclidean domains using diffusion wavelets, while preserving a notion of stability with respect to metric changes in the domain, measured with diffusion maps. The resulting representation is stable to metric perturbations of the domain while being able to capture "high-frequency" information, akin to the Euclidean Scattering.

arXiv commentSubmitted to the International Conference on Learning Representations (ICLR 2019)

Cited by in corpus (15)

• Geometric Deep Learning: Grids, Groups, Graphs, Geodesics, and Gauges
• Stability Properties of Graph Neural Networks
• Invariance-Preserving Localized Activation Functions for Graph Neural Networks
• Stability of Graph Scattering Transforms
• Coarse Graining of Data via Inhomogeneous Diffusion Condensation
• Geometric Scattering Attention Networks
• Scattering Networks on the Sphere for Scalable and Rotationally Equivariant Spherical CNNs
• Geometric Wavelet Scattering Networks on Compact Riemannian Manifolds
• Human Action Recognition with Multi-Laplacian Graph Convolutional Networks
• Stability of Graph Convolutional Neural Networks to Stochastic Perturbations
• Data-Driven Learning of Geometric Scattering Networks
• Uncovering the Folding Landscape of RNA Secondary Structure with Deep Graph Embeddings
• A Block-based Generative Model for Attributed Networks Embedding
• Spatio-Temporal Graph Scattering Transform
• Central and Non-central Limit Theorems arising from the Scattering Transform and its Neural Activation Generalization