PhD Position in Geometric Deep Learning and Graph Signal Processing

ELLIS; France; Sur place; EUR 20 000 - 27 000; Plein temps

Résumé du poste

ELLIS at Télécom Paris/SudParis invites applications for a fully funded PhD position on Signal-aware Graph Summarization with GNN Guarantees. The project spans graph signal processing, graph ML, and geometric deep learning, aiming to preserve information for reconstruction and downstream tasks.

The PhD combines mathematical foundations with modern GNNs, focusing on guarantees, interpretability, and scalability to large graphs, over a 3-year duration.

Qualifications

• Candidate has M2 or equivalent in a related field.
• Strong background in signal processing and machine learning.
• Proficient in Python and PyTorch.

Responsabilités

• Develop methods for signal-aware graph summarization with guarantees.
• Study reconstruction and learning guarantees for graph models.
• Publish results and contribute to collaborations at LTCl and SAMOVAR.

Connaissances

Signal processing

Machine learning

Python

PyTorch

Graph ML

Geometric deep learning

Communication skills

Formation

M2 or equivalent degree

Outils

Python

PyTorch

Description du poste

PhD Position in Geometric Deep Learning and Graph Signal Processing

We are opening a fully funded PhD position at Télécom Paris and Télécom SudParis, Institut Polytechnique de Paris, within the Hi! PARIS SIGMA project, on the topic:

Signal-aware Graph Summarization with GNN Guarantees

This PhD lies at the intersection of graph signal processing, graph machine learning, and geometric deep learning. The goal is to develop principled methods for jointly summarizing graph structures and their associated signals while preserving the information required for signal reconstruction and downstream learning tasks.

The project will combine mathematical foundations in graph signal processing with modern graph neural networks, with a particular focus on reconstruction, learning guarantees, interpretability, and scalability to large graphs.

We are looking for candidates who:

• Hold, or are completing, an M2 or equivalent degree in engineering, data science, computer science, applied mathematics, signal processing, statistics, or a related field
• Have a strong background in signal processing and machine learning
• Have strong programming skills in Python, including PyTorch
• Have a genuine interest in graph signal processing and geometric deep learning
• Are strongly interested in understanding the mathematical foundations behind these areas
• Have good communication skills

Practical information:

• Research teams: MM Team at LTCI and ARMEDIA Team at SAMOVAR
• Duration: 3 years

Full PhD description: https://docs.google.com/viewer?url=raw.githubusercontent.com/jhonygiraldo/jhonygiraldo.github.io/master/assets/pdf/PhD_Position_SIGMA.pdf