GIMC SIMAI Young 2026

Session

MS02.1 - Advances in Neural Network Approximation and Surrogate Modeling for Scientific Machine Learning

4 Jun 2026, 14:00

Pisa

9. NeuberNet: Leveraging Neural Operators for Elasto-Plastic PDE Solutions at Reentrant Corners via Low-Fidelity Elastic Inputs

Tommaso Grossi (TeCIP Institute, Scuola Superiore Sant'Anna)

04/06/2026, 14:00

MS02 - Advances in Neural Network Approximation and Surrogate Modeling for Scientific Machine Learning

Accurately simulating localized plastic strain at geometric discontinuities, such as reentrant corners, remains a significant hurdle due to the computational demands of fully nonlinear modeling. While simplified elastic methods are faster, they fail to account for critical nonlinear effects. To bridge this gap, we introduce NeuberNet, a Multi-Task Nonlinear Manifold Decoder designed to map...

17. Accelerated inverse modeling of tumor evolution via Latent Dynamic Networks

Mikel Mendibe (University of the Basque Country / Tecnalia)

04/06/2026, 14:15

MS02 - Advances in Neural Network Approximation and Surrogate Modeling for Scientific Machine Learning

The characterization of tumor evolution through partial differential equations (PDEs), ranging from reaction-diffusion systems to moving-interface models, provides essential insights into cancer progression. However, utilizing these high-dimensional frameworks in inverse settings to recover patient-specific biophysical properties is often computationally prohibitive due to the requirement for...

33. Data-Driven Resolution Enhancement of Electrostatic Potentials Using NextGenPB

Vincenzo Di Florio (MOX Laboratory, Department of Mathematics, Politecnico di Milano, Piazza Leonardo Da Vinci, 32, Milano, 20133, Italy)

04/06/2026, 14:30

MS02 - Advances in Neural Network Approximation and Surrogate Modeling for Scientific Machine Learning

Electrostatic interactions play a fundamental role in biomolecular recognition, binding affinity, conformational stability, and enzymatic activity. These effects are commonly modeled using the linearized Poisson–Boltzmann equation (LPBE), which provides a continuum description of electrostatics in implicit solvent environments. However, obtaining systematically converged high-resolution...

48. Dynamical reduced order approximation of Wasserstein gradient flows

Isabella Carla Gonnella (SISSA)

04/06/2026, 14:45

MS02 - Advances in Neural Network Approximation and Surrogate Modeling for Scientific Machine Learning

This work introduces a novel dynamical reduced-order approximation framework for Wasserstein gradient flows that leverages the geometric structure of the solution manifold to construct an adaptive low-dimensional representation. The proposed method evolves the solution parametrization through appropriately designed systems of ordinary differential equations, allowing the approximation space...

59. Efficient Bayesian Inference for Uncertainty-Aware Scientific Machine Learning

Dario Coscia (SISSA)

04/06/2026, 15:00

MS02 - Advances in Neural Network Approximation and Surrogate Modeling for Scientific Machine Learning

Surrogate models such as PDE learners and machine learning force fields have become essential tools for approximating complex physical systems at a fraction of the computational cost of high-fidelity solvers. As these models are increasingly deployed in scientific and engineering workflows, quantifying their predictive uncertainty is critical for ensuring reliability and informing data...

61. StabOp: A Data-Driven Stabilization Operator for Reduced Order Modeling

Anna Ivagnes (SISSA)

04/06/2026, 15:15

MS02 - Advances in Neural Network Approximation and Surrogate Modeling for Scientific Machine Learning

Spatial filtering has been widely used in under-resolved simulations of convection-dominated flows and, more recently, as a stabilization strategy in reduced order models (ROMs). However, spatial filters have key unresolved issues, such as determining the best-suited filter for specific applications or choosing an appropriate value for the filter radius.
To address these challenges, we...

63. Continuous Neural Network Approximations of Polygonal Basis Functions

Gioana Teora (Politecnico di Torino)

04/06/2026, 15:30

MS02 - Advances in Neural Network Approximation and Surrogate Modeling for Scientific Machine Learning

We introduce two different neural-network-based strategies for the construction of conforming approximation spaces on general polygonal meshes for the numerical solution of partial differential equations. The proposed methodologies build upon the Virtual Element Method (VEM) paradigm, but replace the implicit definition of local basis functions with explicit neural representations.

In the...

70. Graph Neural Networks for Polytopal Mesh Agglomeration with Applications to Neurodegenerative Processes

Mattia Corti (MOX Laboratory, Department of Mathematics, Politecnico di Milano)

04/06/2026, 15:45

MS02 - Advances in Neural Network Approximation and Surrogate Modeling for Scientific Machine Learning

Agglomeration techniques for polytopal meshes play a key role in reducing the computational cost of large-scale simulations, especially when high-order methods or complex geometries are involved. We introduce a geometrical deep learning framework for automatic mesh agglomeration, in which a Graph Neural Network learns to partition the connectivity graph of three-dimensional meshes and to...