Youtor
DDPS “Infinite Dimensional Optimization for Scientific Machine Learning”
>> YOUR LINK HERE: ___ http://youtube.com/watch?v=JmRMShKxeio
DDPS Talk date: October 25th, 2024 • Speaker: Marius Zeinhofer (University Hospital Freiburg, https://www.linkedin.com/in/marius-ze...) • Description: This talk provides an infinite-dimensional perspective on optimization problems encountered in scientific machine learning and advocates for the paradigm first optimize, then discretize for their solution. This amounts to first choosing an appropriate infinite-dimensional algorithm which is subsequently discretized in the tangent space of the neural network ansatz. To illustrate this point, we show that recently proposed state-of-the-art algorithms for scientific machine learning applications can be derived within this framework. Furthermore, we discuss how to transfer the well-known Kronecker-factored approximation for natural gradient descent to the scientific machine learning setting to allow large-scale applications. • Bio: Dr. Marius Zeinhofer received his Ph.D. from the University of Freiburg, Germany, where he researched multiphysics models for bone growth under the supervision of Prof. Patrick Dondl. He then served as a Postdoctoral Research Fellow at Simula Research Laboratory in Oslo, Norway, focusing on scientific machine learning (SciML) with Prof. Kent-Andre Mardal. Currently, he is a researcher at the University Hospital Freiburg, specializing in AI for healthcare. His ongoing research interests include error analysis and optimization of neural network-based methods for solving partial differential equations (PDEs). • DDPS webinar: https://www.librom.net/ddps.html • 💻 LLNL News: https://www.llnl.gov/news • 📲 Instagram: / livermore_lab • 🤳 Facebook: / livermore.lab • 🐤 Twitter: / livermore_lab • • About LLNL: Lawrence Livermore National Laboratory has a mission of strengthening the United States’ security through development and application of world-class science and technology to: 1) enhance the nation’s defense, 2) reduce the global threat from terrorism and weapons of mass destruction, and 3) respond with vision, quality, integrity and technical excellence to scientific issues of national importance. Learn more about LLNL: https://www.llnl.gov/. • IM release number is: LLNL-VIDEO-871103
#############################
