Graph Neural Networks: an introduction
Автор: Normastic
Загружено: 2021-07-13
Просмотров: 178
An introductive tutorial to Graph Neural Networks, Benoit Gaüzère (LITIS), in close collaboration with Sébastien Adam, Pierre Héroux, Muhammet Balcilar, Guillaume Renton, Paul Honeine and Clément Gledel. (Normastic workshop, feb 2021)
Abstract: This tutorial aims to introduce the problematics and basics of Graph Neural Networks to people unfamiliar with them. After expliciting the difficulties of transposing the sucesses of CNN to graphs, we will present the general message passing framework. This general framework will help to understand the fundamentals of convolutions applied to graphs and their limits. We will also discuss about the standard graph datasets emaning from the recent hype for GNN and the problem of graph generation.
slides: http://www.normastic.fr/wp-content/up...
Доступные форматы для скачивания:
Скачать видео mp4
-
Информация по загрузке:
