Welcome

The detailed program is here.

The registration system is now available at https://scml2025.award-con.com

Due to requests from several authors, the deadline for the submission of papers has been changed to January 15, 2025.

In recent years, machine learning methods for scientific computing have attracted much attention. Many methods are a combination of machine learning and/or theories of physics and/or computational mathematics.

This conference aims to showcase the latest research in these areas, which have been fragmented while pursuing research in the same direction, to bridge the gap between them, and to promote collaboration.

Topics will include, but not limited to

• ML for scientific computing
• ML for model discovery
• Physics-informed neural networks
• Operator learning
• Geometric deep learning
• Numerical method for scientific computing
• Discrete mechanics
• Mathematics for ML for science
• Computational algebra for modeling and simulation

This year, this conference will be held in a hybrid format and online participation is possible. All oral talks will be recoreded and made available to participants.In addition, online oral/poster presentations are also allowed.

Synergies of Machine Learning and Numerics will be held in Osaka after SCML2025, offering a convenient opportunity to attend both events.

Please contact the organizers at yaguchi (at) pearl.kobe-u.ac.jp or scml25-committee (at) geom.jp with any questions.

Conference venue

Kyoto Tower Hotel

721-1 Higashishiokoji-cho, Karasuma-dori Shichijo-sagaru, Shimogyo-ku, Kyoto, 600-8216, Japan

Accommodation

There are many hotels around Kyoto Station, and we are pleased to offer two hotel options for participants attending SCML 2025. Please review the details below to select the plan that best suits your needs.

Common Details for Both Options:

• Check-in: 2nd March 2025 / Check-out: 8th March 8 2025 (6 nights). Participants may request minor adjustments to the check-in or check-out dates, though rates may vary accordingly.
• Meals: No meals included

To request a reservation, please complete the corresponding form based on your chosen accommodation. Please note that submitting the form serves as a reservation request and not a confirmation. Your booking will be confirmed once you receive a confirmation email from SCML organizing committee. The number of rooms available for reservation is limited, and bookings will be made on a first-come, first-served basis.

Option 1: Kyoto Tower Hotel (Conference venue)

• Website: https://www.kyoto-towerhotel.jp/en/
• Room Type: One guest per room. Room preferences cannot be accommodated.
• Rate per night: 12,850 yen (including service charge and tax)
• Reservation Request Form: https://forms.gle/FwMo8FUhcLMmWhk96

Option 2: Hotel Hokke Club Kyoto

• Website: https://global.hokke.co.jp/kyoto/en/
• Room Type: Single room.
• Rate per night: 7,900 yen (including service charge and tax)
• Reservation Request Form: https://forms.gle/WC3ZKcbgf9XfnhMu6

If you experience any difficulty completing the form, please email the following information to m-komatsu at kobe-u.ac.jp.

• Full Name (as on Passport)
• Email Address
• Phone Number
• If applicable, any requested changes to the check-in or check-out dates.

Call for papers and submission guidelines

We welcome paper submissions from all related areas with the above topics.

• Submitted papers will be reviewed by the Program Committee. All accepted papers will be made available to conference participants. However, authors of accepted papers can choose whether to include their paper in the conference proceedings and make it public, or keep it non-public (and hence non-archival). If they choose to keep it non-public (and non-archival), the authors retain the right to publish the paper elsewhere.
• Submission of papers that are under review or have been recently published in a conference or a journal is allowed; but, in that case, these papers cannot be included in the conference proceedings.
• Each accepted presentation will be assigned to either an oral presentation or a poster presentation, according to the review reports.
• Submissions should not exceed four pages, excluding references and supplementary materials.
• All submissions must be in the pdf format based on the SCML style file. Please see template.pdf in the SCML style file for other details.

How to submit a paper

Please submit your paper via Microsoft CMT: https://cmt3.research.microsoft.com/SCML2025

Important Dates

• Paper submissions due: January 5January 15, 2025 (AoE)
• Notification to authors: January 12January 26, 2025 (AoE)

Tutorial/keynote/invited speakers

Several keynote/invited/tutorial talks will be scheduled. In particular, we are planning invited talks by early and mid-career researchers, as this research field is still at the beginning stage and promoting young researchers is very important.

This year, we will invite as invited speakers a few students or young researchers (who have obtained a PhD degree within the five years, i.e., on January 1st, 2020 or later) who have published at least one paper related to SciML in top journals or famous conferences such as NeurIPS, ICML, ICLR. Travel and accommodation expenses for those selected as invited speakers will be covered by the conference organizing committee. Those who wish to apply for this opportunity should send the paper accepted for the above conference to scml25-committee (at) geom.jp by December 13, 2024 (AOE). Please understand that we have a limited budget and that we need to review applications.

The current confirmed tutorial/keynote/invited speakers are

• Nikola Kovachki (NVIDIA)
• Taiji Suzuki (University of Tokyo/RIKEN)
• Takashi Furuya (Shimane University)
• Yury Korolev (Bath University)
• Michael Puthawala (South Dakota State University)
• Chu Haoyu (China University of Mining and Technology)
• Keiya Hirashima (The University of Tokyo)
• Alvaro Fernandez Corral (DESY, Universität Hamburg)

Accepted Talks

The accepted presentations are listed below. This year, we received 40 submissions, of which 23 were accepted (the acceptance rate: 23/40 = 57.5%.)

Oral Presentations

• Regression-Based Physics-informed Neural Network (Reg-PINN) for Magnetopause Tracking

  Po-Han Hou (Imperial College London), Sung-Chi Hsieh (University of Leicester)

• Adjoint-Based Online Learning of Baroclinic Turbulence

  Fei Er Yan(Hong Kong University of Science and Technology), Hugo Frezat(Institut de Physique du Globe de Paris), Julien Le Sommer(niversité Grenoble-Alpes), Julian Mak(Hong Kong University of Science and Technology, National Oceanography Centre), and Karl Otness(New York University)

• Reinforcement Learning for Optimal Trade Execution

  Lufan Wang (University of Waterloo), Justin Wan (University of Waterloo)

• Evidential Physics-Informed Neural Networks

  Hai Siong Tan (Gryphon Center for AI and Theoretical Sciences), Kuancheng Wang (Georgia Institute of Technology), Rafe McBeth (University of Pennsylvania)

• Generalized Lie Symmetries in Physics-Informed Neural Operators

  Xiang Wang (NYU), Zakhar Shumaylov (University of Cambridge), Peter Zaika (University of Cambridge), Ferdia Sherry (University of Cambridge), Carola-Bibiane Schonlieb (University of Cambridge)

• Estimating Distributions of Parameters in Nonlinear State Space Models with Stein Variational Markov Chain Monte Carlo Method

  Koshin Hagimoto (Kobe University), Toshiaki Omori (Kobe University)

• Modeling Coupled Systems by Neural Networks through Poisson-Dirac Formulation

  Razmik Khosrovian (Osaka University), Takaharu Yaguchi (Kobe University), Hiroaki Yoshimura (Waseda University), Takashi Matsubara (Hokkaido University)

• A Hybrid Finite Element and Machine Learning Approach to Willmore Flow

  Martin Rumpf (University of Bonn), Josua Sassen (ENS Paris-Saclay), Christoph Smoch (University of Bonn)

• Improving Regional Weather Forecasts with Neural Interpolation

  James Jackaman (NTNU), Oliver Sutton (King's College London)

• Estimation and Updating of Digital Twin Models via Scientific Machine Learning

  Arjit Seth (The University of Texas at Austin), Tan Bui (The University of Texas at Austin)

Poster Presentations

• Heterogeneous Transfer Learning for Efficient Transitions Between Batch and Continuous Pharmaceutical Manufacturing

  Junya Ihira (Kyoto University Graduate School of Informatics), Keita Yaginuma (Formulation Technology Research Laboratories, Daiichi Sankyo Co., Ltd.), Kanta Sato (Formulation Technology Research Laboratories, Daiichi Sankyo Co., Ltd.), Shota Kato (Kyoto University Graduate School of Informatics), Manabu Kano (Kyoto University Graduate School of Informatics)

• Towards a Diffusion-Based Virtual Subject Generator

  Imran Nasim (IBM), Adam Nasim (Merck)

• Efficient constrained optimisation on the equilibration of unstable baroclinic flows: initial result

  Ho Ching Lee (Hong Kong University of Science and Technology), Julian Mak (Hong Kong University of Science and Technology)

• An Application of the Holonomic Gradient Method to the Neural Tangent Kernel

  Akihiro Sakoda (Kobe University), Nobuki Takayama (Kobe University)

• Fine-Tuning MLP-Mixer Architectures For Extreme Weather Event Prediction

  Imran Nasim (IBM), João Lucas de Sousa Almeida (IBM)

• Advancing Structural Vibration Analysis: Implementation of PINNs for Aerospace Applications

  Jainish Solanki (Indian Institute of Technology Kharagpur), Sakshi Patil (Indian Institute of Technology Kharagpur), Mohammed Rabius Sunny (Indian Institute of Technology Kharagpur)

• Energy-consistent Neural Operator Learning

  Yusuke Tanaka (NTT), Takaharu Yaguchi (Kobe University), Tomoharu Iwata (NTT), Naonori Ueda (RIKEN)

• Refinement of the average vector field method for Hamiltonian systems using neural networks

  Chong Shen (Kobe Univeisity), Baige Xu (Kobe University), Elena Celledoni (Norwegian University of Science and Technology), Brynjulf Owren (Norwegian University of Science and Technology), Takaharu Yaguchi (Kobe University)

• Learning Hamiltonian Density Using DeepONet for Modeling Wave Equations

  Baige Xu (Kobe University), Yusuke Tanaka (NTT Corporation), Takashi Matsubara (Hokkaido University), Takaharu Yaguchi (Kobe University )

• Learning Hamiltonian Partial Differential Equations Using DeepONet with a Symplectic Branch Network

  Yeang Makara (Kobe University), Yusuke Tanaka (NTT Communication Science Laboratories), Takashi Matsubara (Faculty of Information Science and Technology), Takaharu Yaguchi (Kobe University)

• An Infinite Dimensional LSSL with Infinite Dimensional HiPPO

  Atsushi Takabatake (Kobe University), Takaharu Yaguchi (Kobe University)

• A Moderate Survey of Sketching Techniques Comparison for Randomized Numerical Linear Algebra under Machine Learning Setting

  Yuqi Liu (University of California, Berkeley), Leon Mikulinsky(University of California, Berkeley), Konstantin Zörner(University of California, Berkeley), James Demmel(University of California, Berkeley)

• PM 2.5 Advection-Diffusion with Multiple Sources and LSTM Neural Network Surrogate Model Optimization

  Kevin Yotongyos(Chiang Mai University), Somchai Sriyab(Chiang Mai University)

Registration

Registration is required for you to participate in the conference. In particular, to present your work, at least one of the authors should make a registration.

The system for registration is available at https://scml2025.award-con.com

Registration fee

• Early registration (until February 3 February 10, 2025)
  • Regular participant: 50000JPY
  • Regular participant (online participation only): 40000JPY
  • Student participant: 30000JPY
  • Student participant (online participation only): 20000JPY
    
• Regular registration
  • Regular participant: 60000JPY
  • Regular participant (online participation only): 40000JPY
  • Student participant: 40000JPY
  • Student participant (online participation only): 20000JPY

Organizers

• Takaharu Yaguchi (Kobe University)
• Hiroaki Yoshimura (Waseda University)
• Nobuki Takayama (Kobe University)
• Toshiaki Omori (Kobe University)
• Takashi Matsubara (Osaka University)
• Kumiko Hori (National Institute for Fusion Science)
• Mizuka Komatsu (Kobe University)
• Baige Xu (Kobe University)