Réseaux de Neurones à Graphes Géométriques - Gif-sur-Yvette, France - Université Paris-Saclay GS Informatique et sciences du numérique

Université Paris-Saclay GS Informatique et sciences du numérique
Université Paris-Saclay GS Informatique et sciences du numérique
Entreprise vérifiée
Gif-sur-Yvette, France

il y a 3 semaines

Sophie Dupont

Posté par:

Sophie Dupont

beBee Recruiter
Description
Réf
-
ABG-118374

ADUM-51667

  • Sujet de Thèse 23/11/2023
  • Université Paris-Saclay GS Informatique et sciences du numérique
  • Lieu de travail
  • Gif sur Yvette
  • France
  • Intitulé du sujet
  • Mots clés
  • Apprentissage automatique, réseaux de neurones en graphes géométriques, apprentissage profond géométrique, quantification des incertitudes
  • Graph representation learning, geometric deep learning, graph neural networks, uncertainty quantification, molecular systems
    Description du sujet:
  • Graph Neural Networks (GNNs) (e.g., [1,2]) currently constitute state-of-the-art models for solving
- prediction tasks on graphs. Through the flexible formulation of the message passing mechanism,

  • GNNs can learn informative latent representations of graph entities at different resolution levels
  • The goal of this Ph.
D. position is to develop uncertainty-aware geometric GNN models that can
- be leveraged in real-world prediction problems, primarily in materials modeling [6]. The objectives
- of the position are the following:

  • Objective 1. First, we aim to develop geometric GNNs that balance predictive performance
and scalability. Following our recent work [7], we will study methodologies to
- enforce symmetries through appropriate data augmentation strategies rather than directly in
- the model architecture.
  • Objective 2. Then, we will investigate uncertainty estimation and quantification mechanisms
in geometric GNNs. This includes Bayesian formulations of GNNs in which the message pass
ing mechanism propagates uncertainty estimates of nodes beyond feature representations.

  • Besides, we are interested in studying a conformal prediction framework for modelagnostic
- uncertainty estimates [8].- in complex molecular systems arising in materials modeling. This will be achieved in collaboration with experts in the broader context of leveraging machine learning techniques to analyze the properties of materials to address energy challenges posed by climate change [9].
The goal of this Ph.
D. position is to develop uncertainty-aware geometric GNN models that can
- be leveraged in real-world prediction problems, primarily in materials modeling [6].

The objectives of the position are the following:

  • Objective 1. First, we aim to develop geometric GNNs that balance predictive performance
and scalability. Following our recent work [7], we will study methodologies to
- enforce symmetries through appropriate data augmentation strategies rather than directly in
- the model architecture.
  • Objective 2. Then, we will investigate uncertainty estimation and quantification mechanisms
in geometric GNNs. This includes Bayesian formulations of GNNs in which the message pass
ing mechanism propagates uncertainty estimates of nodes beyond feature representations.

  • Besides, we are interested in studying a conformal prediction framework for modelagnostic
- uncertainty estimates [8].- in complex molecular systems arising in materials modeling. This will be achieved in collaboration with experts in the broader context of leveraging machine learning techniques to analyze the properties of materials to address energy challenges posed by climate change [9].

Début de la thèse : 01/10/2024
Nature du financement:

Précisions sur le financement:


  • Programme COFUND DeMythif

AI
Présentation établissement et labo d'accueil:

  • Université Paris-Saclay GS Informatique et sciences du numérique

Etablissement délivrant le doctorat:

  • Université Paris-Saclay GS Informatique et sciences du numérique
    Ecole doctorale:
  • 580 Sciences et Technologies de l'Information et de la Communication17/01/2024
Plus d'emplois de Université Paris-Saclay GS Informatique et sciences du numérique
Voir tous les emplois de Université Paris-Saclay GS Informatique et sciences du numérique