Complex Network Classification with Convolutional Neural Network

TL;DR

This paper introduces a novel framework called CNC that combines network embedding and CNN to classify complex networks accurately, automatically extracting features despite their non-Euclidean nature.

Contribution

The paper presents a new network classification framework integrating embedding and CNN, effective on synthetic and real-world data, with high accuracy and automatic feature extraction.

Findings

High classification accuracy on synthetic and real networks

Robustness to network variations

Automatic feature extraction capability

Abstract

Classifying large scale networks into several categories and distinguishing them according to their fine structures is of great importance with several applications in real life. However, most studies of complex networks focus on properties of a single network but seldom on classification, clustering, and comparison between different networks, in which the network is treated as a whole. Due to the non-Euclidean properties of the data, conventional methods can hardly be applied on networks directly. In this paper, we propose a novel framework of complex network classifier (CNC) by integrating network embedding and convolutional neural network to tackle the problem of network classification. By training the classifiers on synthetic complex network data and real international trade network data, we show CNC can not only classify networks in a high accuracy and robustness, it can also extract the features of the networks automatically.