TIEJ5100 COM1: Graph Neural Networks (JSS34) (1 op)

Avaa opintojakson esite Sisussa
Opinnon taso:
Jatko-opinnot
Arviointiasteikko:
Hyväksytty - hylätty
Suorituskieli:
englanti
Vastuuorganisaatio:
Informaatioteknologian tiedekunta
Opetussuunnitelmakaudet:
2025-2026

Kuvaus

Relational learning

  • Overview
  • Graph neural networks (GNNs)
  • Spectral perspective
  • Spatial perspective
  • Main architectures
  • Dealing with oversmoothing and oversquashing
  • Theory of GNNs
  • Expressivity
  • Generalization
  • Extensions
  • High-order GNNs
  • GNNs for knowledge graphs
  • Graph Transformers
  • Hypergraph neural networks
  • Topological neural networks
  • Generative models for graphs

Osaamistavoitteet

Understanding main concepts and methods in relational learning   

Esitietojen kuvaus

Minimum: basic level machine learning courses. Recommended: advanced courses on machine learning and deep learning 

Oppimateriaalit

  • William L. Hamilton, Graph Representation Learning. Synthesis Lectures on AI and ML, Vol. 14, No. 3. 2020.
  • Michael M. Bronstein and Joan Bruna and Taco Cohen and Petar Veličković, Geometric Deep Learning: Grids, Groups, Graphs, Geodesics, and Gauges. ArXiv, 2021.
  • L. Wu, P. Cui, J. Pei, and L. Zhao. Graph Neural Networks: Foundations, Frontiers, and Applications. Springer, Singapore, 2022
  • M. Hajij et. al., Topological Deep Learning: Going Beyond Graph Data. ArXiv, 2023.
  • Stefanie Jegelka. Theory of Graph Neural Networks: Representation and Learning. ArXiv, 2022.

Suoritustavat

Tapa 1

Kuvaus:
Lectures and demonstrations. Each student is required to give a presentation on the final day.
Arviointiperusteet:
Pass/fail
Opetusajankohta:
Periodi 1
Valitaan kaikki merkityt osat
Suoritustapojen osat
x

Osallistuminen opetukseen (1 op)

Tyyppi:
Osallistuminen opetukseen
Arviointiasteikko:
Hyväksytty - hylätty
Arviointiperusteet:
<p>Pass/fail</p>
Suorituskieli:
englanti
Työskentelytavat:

Lectures and demonstrations. Each student is required to give a presentation on the final day.

Oppimateriaalit:
  • William L. Hamilton, Graph Representation Learning. Synthesis Lectures on AI and ML, Vol. 14, No. 3. 2020.
  • Michael M. Bronstein and Joan Bruna and Taco Cohen and Petar Veličković, Geometric Deep Learning: Grids, Groups, Graphs, Geodesics, and Gauges. ArXiv, 2021.
  • L. Wu, P. Cui, J. Pei, and L. Zhao. Graph Neural Networks: Foundations, Frontiers, and Applications. Springer, Singapore, 2022
  • M. Hajij et. al., Topological Deep Learning: Going Beyond Graph Data. ArXiv, 2023.
  • Stefanie Jegelka. Theory of Graph Neural Networks: Representation and Learning. ArXiv, 2022.

Opetus