Geometric Scattering for Graph Data Analysis

TL;DR

This paper extends scattering transforms to graph data, creating features that are stable and informative for tasks like graph classification and data exploration in social networks and biochemistry.

Contribution

It introduces a geometric scattering framework for graphs, demonstrating its stability and utility in graph analysis tasks, bridging deep learning and graph theory.

Findings

Effective graph classification on social network data

Useful for data exploration in biochemistry

Features retain meaningful variability and relations

Abstract

We explore the generalization of scattering transforms from traditional (e.g., image or audio) signals to graph data, analogous to the generalization of ConvNets in geometric deep learning, and the utility of extracted graph features in graph data analysis. In particular, we focus on the capacity of these features to retain informative variability and relations in the data (e.g., between individual graphs, or in aggregate), while relating our construction to previous theoretical results that establish the stability of similar transforms to families of graph deformations. We demonstrate the application the our geometric scattering features in graph classification of social network data, and in data exploration of biochemistry data.