Skip to content
This repository was archived by the owner on Oct 26, 2024. It is now read-only.

future-ai-org/mlnet-complex-networks-py

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

44 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

mlnet: classifying complex graph networks with machine learning


the study of complex networks permeates all of the science
  • we can assign complex networks into four general classes (although there is some overlap between them):
    • technological networks (e.g., Internet, the telephone network, power grids, transportation network)
    • information networks (e.g., the world wide web, citation networks)
    • biological networks (e.g., biochemical network, neural networks, ecological networks)
    • social networks
characterizing complex network's structures is a key to understanding any unifying principles underlying their topology
  • previous works have shown that many topological properties can vary for different types of system, however these works generally focus only a few characteristics at the time
  • in this project, we present the first part of a method to characterize complex networks by performing an extensive analysis of the global and local topological features of networks
  • in a second part, these features are used into input vectors for a SVN classifier, establishing an efficient way of learning the classification of complex networks


the input data and features


  • prior to using this software:
  • vectors can have different normalizations (snowball and metropolis hastings random walk samplings for different sizes)
  • vectors can contain entire graphs for some of the features (that were possible to be calculated)


feature selection and classifiers


  • we perform the classification of the network features using several classifiers:
    • SVM (supervised)
    • logistic regression (supervised)
    • adaboost (supervised)
    • EM (unsupervised)


analysis and plots


  • comparisons of many classifiers and the plots are available under each classifier's folder.


installation


virtualenv .venv
source .venv/bin/activate
pip install -r requirements.txt

About

👾 classifying complex graph networks with machine learning

Topics

Resources

Stars

Watchers

Forks

Contributors