v255: Proceedings of ML Meets Differential Equations

2026-06-10 · Source: Proceedings of Machine Learning Research · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Mathematics & Computational Sciences, Engineering & Applied Sciences · Depth: Expert, quick

Summary

The 1st ECAI Workshop on "Machine Learning Meets Differential Equations: From Theory to Applications," held on October 20, 2024, in Santiago de Compostela, Spain, presented Volume 255 of its proceedings. This volume features nine research papers exploring the intersection of machine learning and various differential equation paradigms. Key contributions include optimizing Neural Fractional Differential Equations, applying neural ensembles for vibrating beam system identification, and developing Time and State Dependent Neural Delay Differential Equations. Further research delves into accelerating Hopfield Network dynamics and utilizing Neural Ordinary Differential Equations for 2D flow analysis in hydraulic structures and optimal control of coastal ecosystems. The workshop also introduces PINNtegrate for Physics-Informed Neural Network-based integral-learning, addresses diffusion model likelihood in conditional settings, and presents methods for extracting sparse models from implicit dynamics.

Key takeaway

This workshop presents cutting-edge research integrating machine learning, particularly neural networks, with various differential equations (ODEs, FDEs, DDEs) for complex system modeling and analysis. Papers detail novel techniques for optimizing neural DEs, accelerating network dynamics, and applying PINNs and NODEs to problems ranging from hydraulic flow analysis and vibrating beam identification to coastal ecosystem control. These advancements offer critical insights for professionals in scientific computing and AI-driven engineering seeking enhanced predictive modeling, system identification, and optimal control solutions.

Topics

• Neural Differential Equations
• Physics-Informed Neural Networks
• Hopfield Networks
• Diffusion Models
• System Identification

Best for: AI Scientist, Research Scientist

Related on AIssential

• v277: Proceedings of ML-DE 2025 — The workshop explores diverse applications of machine learning to differential equations across various scientific and engineering domains.
• Toward AI-Driven Digital Twins for Metropolitan Floods: A Conditional Latent Dynamics Network Surrogate of the Shallow Water Equations — CLDNet significantly accelerates metropolitan flood forecasting by using a latent neural ODE and coordinate-based decoding.
• Robust Parameter and State Estimation in Multiscale Neuronal Systems Using Physics-Informed Neural Networks — PINNs offer robust parameter and state estimation in multiscale neuronal systems, overcoming limitations of traditional methods.
• A Novel Tensor Product-Based Neural Network for Solving Partial Differential Equations — TPNet solves PDEs and approximates functions via a tensor-product basis and direct least-squares, bypassing gradient training.
• Physics-Guided Dimension Reduction for Simulation-Free Operator Learning of Stiff Differential--Algebraic Systems — An extended Newton implicit layer enables simulation-free, accurate operator learning for stiff DAEs by enforcing constraints and reducing dimensions.
• Physics-Informed Neural Networks: A Didactic Derivation of the Complete Training Cycle — This guide demystifies PINN training by exposing the full algebraic cycle, from forward pass to gradient descent.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Proceedings of Machine Learning Research.