Happy to be selected to the list of 2022 Top 10 Outstanding Chinese American Youth, during Asian American and Pacific Islander Heritage Month.
Author: Jian-Xun Wang
Invited Talk at USACM UQ Virtual Seminar Series
I will give invited talk at #USACM#UQ Virtual Seminar Organized by USACM UQ TTA. (Time: 15:00 PM-16:00 PM (EST US), 5/12/2022. Zoom Link: listed below). The topic is about “how to leverage physics in ML for computational mechanics — Physics-informed, PDE-structure preserved Learning for problems with complex geometries”. I am happy to have Prof. Michael Brenner be my discussant sharing his great insights in the field of scientific machine learning #SciML. Also thank the organization committee: Abani Patra, Serge Prudhomme, Johann Guilleminot.
In this talk, I will discuss our recent developments of on SciML from several different aspects, including (1) how to leverage physics to inspire network architecture design (PDE-preserved deep learning #PPNN), (2) use physics to inform network training #PINN, (3) physics-informed geometric deep learning (#GeometricDL) for complex geometries and irregular domains. This work is supported National Science Foundation (NSF)#OAC#CMMI, Aerospace & Mechanical Engineering at Notre Dame, Center for Sustainable Energy at Notre Dame, ND-ECI, ND Lucy Institute of Data Science.
Welcome to join us at https://lnkd.in/gH6AHp3V. See you on Thursday
Here is the talk recording
Will host a MS a SIAM UQ22 and our group will give 5 presentations
![](https://sites.nd.edu/jianxun-wang/files/2022/03/image-1024x277.png)
Together with Yeonjong Shin (Brown University) and Xueyu Zhu (U Iowa), we will co-host a 3-part mini-sympoiusm entitled “Recent Advances in Machine Learning and Data-Driven Methods for Physical Sciences and Engineering”, where 12 speakers will give talks on scientific machine learning and data-driven modeling.
- 9:30 AM – 11:30 AM, 04/12/2022, (Augusta E – 7th Floor) MS2 Part I
- 8:10 AM – 10:10 AM, 04/13/2022, (Augusta E – 7th Flow) MS45 Part II
- 8:10 AM – 10:10 AM, 04/14/2022, (Augusta E – 7th Flow) MS89 Part III
Our group will give 5 talks at SIAM UQ22.
- 3:00 – 3:25 PM, 04/12/2022, (Chastain J – 6th Floor), MS26, Xinyang Liu et al. Physics-Informed Model-Based Reinforcement Learning with Quantified Uncertainties.
- 3:15 – 3:40 PM, 04/13/2022, (Peachtree 2 – 8th Floor), MS71, Luning Sun et al. Deep Bayesian Spline Learning for Closed-Form Equation Discovery of Nonlinear Dynamics with Quantified Uncertainty
- 3:15 – 3:40 PM, 04/13/2022, (Peachree A 8th Floor), MS72, Han Gao et al. A long-span geometric deep learning framework based on attention mechanism for fast surrogate predictions of spatiotemporal dynamics
- 8:40 – 9:05 AM, 04/14/2022, (Augusta E – 7th Floor), MS89, Xinyang Liu et al. PDE-preserved network architecture for predicting spatiotemporal dynamics
- 5:30 – 5:55 PM, 04/14/2022, (Chastain E – 6th Floor), MS123, Pan Du et al. Fast Surrogate of 3-D Patient-Specific Computational Fluid Dynamics Using Statistical Shape Modeling and Deep Learning
Our paper has been accepted by ICLR 2022
Happy to share our work “Predicting Physics in Mesh-Reduced Space with Temporal Attention”, which has been accepted to ICLR 2022. This is a collaborative work with Tufts and DeepMind. Particularly congrats to the student authors X. Han and H. Gao. Arxiv link: https://lnkd.in/dxvZRXWC.
![](http://sites.nd.edu/jianxun-wang/files/2022/02/ICLRAnimation_medium.gif)
Highest Citation Paper Award
Our article “L. Sun, J.-X. Wang, Physics-constrained bayesian neural network for fluid flow reconstruction with sparse and noisy data, Theor. App. Mech. Let. 10 (2020) 161-169 ” winning the 2021 Highest Citation Paper Award of Theoretical & Applied Mechanics Letters on January 2022. Congrats to Luning!
![](https://sites.nd.edu/jianxun-wang/files/2022/01/6-Highest-Citation-Sun-LuningWang-Jianxun-1024x730.jpg)
Will give a seminar talk @ AI institute at Renmin University of China
![](https://sites.nd.edu/jianxun-wang/files/2021/11/640-641x1024.png)
Give a seminar talk @ ANL
![](https://sites.nd.edu/jianxun-wang/files/2021/11/image-1024x895.png)
Pan Du won poster award @ Fall Symposium of Lucy Institute of Data Science and Society. Congrats!
Pan Du won poster award @ Fall Symposium of Lucy Institute of Data Science and Society. Congrats! The Mayor of South Bend presented the award.
![](https://sites.nd.edu/jianxun-wang/files/2021/11/image-1024x768.jpeg)
Will host a MS at USNCCM16 and our group will give 3 presentations
![](https://sites.nd.edu/jianxun-wang/files/2021/07/image-4-1024x326.png)
Together with Danial, Kathryn, Alireza, and Hao, we will co-host a mini-sympoisum entiled Physics-Informed Learning and Data-Enabled Predictive Modeling and Discovery of Complex Systems, where 15 speakers will give excellent talks on scientific machine learning and data-driven predictive modeling. (10:00-18:00 (UTC), 07/26/2021). Please check out.
Our group will also give 3 talks at USNCCM:
- 14:00-14:20 (UTC) 07/27/2021: X.-Y. Liu* and J.-X. Wang, Physics-informed model-based deep reinforcement learning for dynamic control.
- 17:20-17:40 (UTC) 07/27/2021: P. Du*, X. Zhu, J.-X. Wang, Surrogate modeling for 3-D patient-specific hemodynamics using statistical shape modeling and deep learning
- 15:00-15:20 (UTC) 07/27/2021: J.-X. Wang*, H. Gao, L. Sun, Physics-informed discretization-based learning: a unified framework for solving PDE-constrained forward and inverse problems
Will host a mini-symposium at SIAM AN 21 (July 19-23), and Xinyang will give a presentation on MBRL for control.
![](https://sites.nd.edu/jianxun-wang/files/2021/07/image-3.png)
Together with Prof. Huan Xun@University of Michigan, we will host a two-section mini-symposium (MS48) entiled: Physics-aware machine learning for solving and discovering PDEs, part I (MS48) and part II (MS105)
Part I (MS 48), Tuesday, July 20
- 4:30-4:55 Deep Neural Network Modeling of Unknown PDEs in Nodal Space abstract Zhen Chen, Ohio State University, U.S.;
Victor Churchill, Dartmouth College, U.S.; Kailiang Wu and Dongbin Xiu, Ohio State University, U.S.
- 5:00-5:25 Deep Learning Methods for Discovering Physics from Data abstract Joseph Bakarji, Jared L. Callaham, and Kathleen Champion, University of Washington, U.S.; J. Nathan Kutz, University of Washington, Seattle, U.S.; Steve Brunton, University of Washington, U.S.
- 5:30-5:55 Data-Driven Learning of Nonlocal Models:from High-Fidelity Simulations to Constitutive Laws abstract Yue Yu and Huaiqian You, Lehigh University, U.S.; Stewart Silling and Marta D’Elia, Sandia National Laboratories, U.S.
- 6:00-6:25 Physics-informed Dyna-Style Model-Based Deep Reinforcement Learning for Dynamic Control abstract
Xinyang Liu and Jianxun Wang, University of Notre Dame, U.S.
Part II (MS 105), Friday, July 23
- 3:30-3:55 Optimal Experimental Design for Variational System Identification of Material Physics Phenomena abstract Wanggang Shen, Zhenlin Wang, Krishna Garikipati, and
Xun Huan, University of Michigan, U.S.
- 4:00-4:25 Learning Stochastic Closures Using Sparsity-Promoting Ensemble Kalman InversionabstractJinlong Wu, Tapio Schneider, and Andrew Stuart, California Institute of Technology, U.S.
- 4:30-4:55 PhyCRNet: Physics-Informed Convolutional-Recurrent Network for Solving Spatiotemporal PDEsabstract
Pu Ren and Chengping Rao, Northeastern University, U.S.; Jianxun Wang, University of Notre Dame, U.S.; Yang Liu and Hao Sun, Northeastern University, U.S.
- 5:00-5:25 Practical Uncertainty Quantification for Learning Partial Differential Equations from Data with Deep EnsemblesabstractSteven Atkinson and Panagiotis Tsilifis, GE Global Research, U.S.