REMODEL-DSC Workshop on Machine Learning and Physics
Hokkaido University and Online (Hybrid), Japan.
August 30 to September 2, 2024
Hokkaido University and Online (Hybrid), Japan.
August 30 to September 2, 2024
In recent years, scientific machine learning, which is combinations of machine learning and scientific computing, has attracted much attention. This workshop aims to introduce the latest results in this field and to provide opportunities to meet, interact and start collaboration.
This workshop is supported by Horizon Europe MSCA Staff Exchanges: Research Exchanges in the Mathematics of Deep Learning with Applications (REMODEL,) JST ASPIRE: Deep Scientific Computing: integration of physical structure and deep learning through mathematical science (DSC) and JST CREST Mathematical Information Platform: Structure Preserving System Modeling and Simulation Basis Based on Geometric Discrete Mechanics.
14:00-14:50 Yusuke Tanaka (NTT): Gaussian Processes for Spatio-temporal Data Analysis
15:00-15:50 Linyu Peng (Keio University): Moving frames, invariant variational problems and invariant variational integration
16:00-16:50 Takaharu Yaguchi (Kobe University): Structure-preserving methods for a class of dissipative differential equations
10:00-12:00 Poster Presentations
Baige Xu, Yusuke Tanaka, Takashi Matsubara, Takaharu Yaguchi: Application of DeepONet for learning Hamiltonian PDEs
Atsushi Takabatake, Baige Xu, Takaharu Yaguchi: Hyperbolic Partial Differential Equations Derived From Hippo Matrices
Razmik Arman Khosrovian, Takaharu Yaguchi, Takashi Matsubara: Learning the Dynamics and Connectivity of Coupled Systems via Port-Hamiltonian Neural Networks
13:00-15:00 Discussions on Open Problems
10:00-12:00 Discussions on Accelerating International Collaboration
13:00- Free Afternoon
10:00-10:50 Takashi Matsubara (Hokkaido University): Content creation with geometric inductive bias
11:00-11:50 Brynjulf Owren (NTNU): Neural network architectures based on non-expansive layer maps on Euclidean and Riemannian spaces
13:00-13:50 Elena Celledoni (NTNU): Deep learning of diffeomorphisms with applications
14:00-14:50 Daisuke Inoue (Toyota Central R&D Labs): An uncertainty-aware, mesh-free numerical method for Kolmogorov PDEs
15:00-15:50 Yuka Hashimoto (NTT): Reproducing kernel Hilbert C*-module for data analysis
Clark Memorial Student Center, Seminar Room 2 (3rd floor), Hokkaido University.
Kita 8, Nishi 5, Kita-ku, Sapporo, Hokkaido, 060-0808, Japan
Takaharu Yaguchi
Takashi Matsubara
Masaaki Imaizumi
Yusuke Tanaka
Mizuka Komatsu
Baige Xu