加入支持让我们有继续维护的动力！会员畅享查看所有预告
立即购买
Axis-augmented physics-informed neural network for Poisson-Boltzmann equation
- 来源：
- 学校官网
- 收录时间：
- 2025-10-18 15:47:04
- 时间：
- 2025-10-17 10:00:00
- 地点：
- 2号学院楼220
- 报告人：
- Gwanghyun Jo
- 学校：
- -/-
- 关键词：
- physics-informed neural network, PINN, Poisson-Boltzmann equation, interface problem, Sobolev extension, numerical methods, deep learning
- 简介：
- Despite the success of physics-informed neural networks (PINNs) in various partial differential equations, their application to interface problems remains limited. This talk presents an axis-augmented PINN that reformulates elliptic interface problems as continuous problems in a higher-dimensional space using Sobolev extension techniques, enabling effective treatment of elliptic interface problems with discontinuous solutions. Numerical results for the Poisson-Boltzmann equation are presented. Additionally, a convergence analysis of the axis-augmented PINN applied to the Poisson-Boltzmann equation is provided.
- -/- 11
报告介绍：
Despite the success of physics-informed neural networks (PINNs) in various partial differential equations, their application to interface problems remains limited. This talk presents an axis-augmented PINN that reformulates elliptic interface problems as continuous problems in a higher-dimensional space using Sobolev extension techniques, enabling effective treatment of elliptic interface problems with discontinuous solutions. Numerical results for the Poisson-Boltzmann equation are presented. Additionally, a convergence analysis of the axis-augmented PINN applied to the Poisson-Boltzmann equation is provided.
报告人介绍：
Gwanghyun Jo is currently an Associate Professor in the Department of Mathematics and Data Science at Hanyang University and serves as an external director at Deltoid Inc. His research focuses on numerical artificial intelligence, including the Virtual Element Method, Conditioned Generation, Deep Reinforcement Learning, and Mathematical Modeling, with over 30 SCI-indexed publications. He received his B.S., M.S., and Ph.D. degrees from KAIST in 2011, 2013, and 2018, respectively. From March 2018, he conducted postdoctoral research at the Department of Mathematical Sciences, KAIST, and was promoted to Assistant Professor in September 2019. He served as an Associate Professor at Kunsan National University from October 2022 and joined Hanyang University in September 2023. He is also the Editor-in-Chief of the Journal of Information and Communication Convergence Engineering and has received awards including the Outstanding Editor Award from the Korean Institute of Information Scientists and Engineers (2024) and the Best Paper Award at 2023 ICFICE.
购买下会员支持下吧...用爱发电已经很久了 立即购买
