ATTENTION/WARNING - NE PAS DÉPOSER ICI/DO NOT SUBMIT HERE

Ceci est la version de TEST de DIAL.mem. Veuillez ne pas soumettre votre mémoire sur ce site mais bien à l'URL suivante: 'https://thesis.dial.uclouvain.be'.
This is the TEST version of DIAL.mem. Please use the following URL to submit your master thesis: 'https://thesis.dial.uclouvain.be'.
 

A Comparison of Machine Learning Algorithms for Graph Semi-Supervised Classification

Feuillen, Thomas
(2020)

Files

FEUILLEN_59931000_2020.pdf
  • Closed access
  • Adobe PDF
  • 1.36 MB

Details

Supervisors
Saerens, Marco
Faculty
Louvain School of Management
Degree label
Master [60] en sciences de gestion (horaire décalé)
Abstract
This work tackles the challenge of learning from partly labelled graph data in the context of node classification. To that end, two families of algorithms are studied : the Support Vector Machine and Convolutional Neural Network. These methods are not directly fitted to deal with the structured information contained in graph data or a network of nodes with features. Thus, we study modified versions of these algorithms that can leverage the full richness of the studied datasets. The algorithms are the Auto-SVM and the Graph Convolutional Network. These will be implemented and evaluated using Python libraries and Google Colab. Simulations on publicly available datasets are provided to showcase the comparison between the different methods before concluding.