Affordable Access

Access to the full text

Deep Convolutional Neural Networks for Feature Extraction of Images Generated from Complex Networks Topologies

Authors
  • Xu, Ye
  • Chi, Yun
  • Tian, Ye
Type
Published Article
Journal
Wireless Personal Communications
Publisher
Springer-Verlag
Publication Date
Feb 06, 2018
Volume
103
Issue
1
Pages
327–338
Identifiers
DOI: 10.1007/s11277-018-5445-7
Source
Springer Nature
Keywords
License
Yellow

Abstract

To identify topology features of different complex network topology is essential in network science researches. Apart from traditional tools in doing such jobs such as power-law, a proved method of convolutional neural network (CNN) is introduced into this research field after we re-format the complex network topology adjacent matrix into an image. We design a CNN of overall 10 layers comprising convolutional layers, pooling layers and a softmax dense layer at last to extract relevant features and classify such features. Experiments show that the CNN models can effectively extract target features and result in an average accuracy rate of 95.65% in feature classification.

Report this publication

Statistics

Seen <100 times