Posters

This page contains a collection of posters presented by me and members of my research group. They could serve as a good introduction/overview to some of the research topics we work on.

Poster presenters are highlighted in boldface.

Click on a poster's title or thumbnail to download its PDF.

Finite-Data Error Bounds for Approximating the Koopman Operator

D. Fassler, J. Bramburger, and S. Brugiapaglia
• CRM Workshop "Approximation and Learning in High Dimensions", 2025
• SIAM/CAIMS Annual Meeting 2025

Deep greedy unfolding: Sorting out argsorting in greedy sparse recovery algorithms
S. M.-Taheri, M. Colbrook, and S. Brugiapaglia

• CRM Workshop "Approximation and Learning in High Dimensions", 2025
• SIAM/CAIMS Annual Meeting 2025

Deep Learning Approximation of Matrix Functions: FromFeedforward Neural Networks to Transformers
R. Padmanabhan andS. Brugiapaglia
• 2025 CMS Summer Meeting🏆 CMS President Award

• CRM Workshop "Approximation and Learning in High Dimensions", 2025

Sparse polynomial approximation of diffusion on community-structured graphs
G. A. D'Inverno, K. Ajavon, and S. Brugiapaglia
• Graph Signal Processing Workshop 2025
• CRM Workshop "Approximation and Learning in High Dimensions", 2025
• YAMC 2025

Practical existence theorems for deep learning in high dimensions
S. Brugiapaglia
• BIRS Workshop “Uncertainty Quantification in Neural Network Models”, 2025

Can transformers approximate functions of matrices?
R. Padmanabhan and S. Brugiapaglia
• 2024 CMS Summer Meeting

Algorithmic auditing of platform real estate and Montreal's housing crisis
S. Brugiapaglia, R. Chodkowski, S. Grant, S. Ishkhanian, A. Renzi, and T. Vukov
• Interdisciplinary AI Forum, Concordia, 2024

Greedy deep unfolding: New neural networks based on sparse recovery algorithms
S. M.-Taheri, M. Colbrook, and S. Brugiapaglia
• 2023 CMS Winter Meeting

Motion detection using dynamic mode decomposition
M. Mignacca, S. Brugiapaglia, and J. Bramburger
• 2023 CMS Winter Meeting🏆 CMS Student Committee Award

Surrogate models for diffusion on graphs: a high-dimensional polynomial approach
K. Ajavon, S. Brugiapaglia, and P. Gora
• 2023 CMS Winter Meeting

LASSO-inspired variants of weighted orthogonal matching purstuit with applications to high-dimensional approximation
S. M.-Taheri and S. Brugiapaglia
• 2021 CANSSI Showcase

Generalizing Outside the Training Set: When Can Neural Networks Learn Identity Effects?
S. Brugiapaglia, M. Liu, and P. Tupper
• CogSci 2020

Google Sites

Report abuse