One Node at a Time: Node-Level Network Classification

TL;DR

This paper shows that individual nodes in networks carry enough structural information to classify the entire network, enabling new applications like small sample network classification and bootstrapping.

Contribution

It establishes the link between node-level features and network classification, demonstrating that nodes reflect the overall network's structural patterns.

Findings

Node features can predict network categories accurately.

Node-level classification enables whole-network categorization from small samples.

The approach applies to various network datasets and models.

Abstract

Network classification aims to group networks (or graphs) into distinct categories based on their structure. We study the connection between classification of a network and of its constituent nodes, and whether nodes from networks in different groups are distinguishable based on structural node characteristics such as centrality and clustering coefficient. We demonstrate, using various network datasets and random network models, that a classifier can be trained to accurately predict the network category of a given node (without seeing the whole network), implying that complex networks display distinct structural patterns even at the node level. Finally, we discuss two applications of node-level network classification: (i) whole-network classification from small samples of nodes, and (ii) network bootstrapping.