Description: This is a 6-week course meeting twice a week as one lecture and one lab. This course will provide an overview of PIML techniques including: soft penalty constraints, neural ODE, universal differential equations (UDE), and PDE super-resolution. During the course, we will make use of several ML frameworks including TensorFlow, PyTorch, Julia, and JAX that make use of GPUs on the HPC.
Presenter(s): Dr. Wes Brewer, Dr. Mathew Boyer, & Dr. Joel Bretheim - PET, GDIT
Location: Webcast
Date & Time: September 8 - October 15, 2020 (Tuesday & Thursday), 4:00p - 5:30p ET
Prerequisites: Attendees should have a basic knowledge of Python programming and using Numpy libraries on the HPCs (See Introduction to Numerical Python (NumPy) ). Attendees should have HPC access in order to use GPUs nodes. Users may connect to the HPC using iLauncher (See Data Analytics Quick Start with iLauncher ). Knowledge of differential equations will be helpful for the course, but not required.
Distribution Statement D. Distribution authorized to the Department of Defense and U.S. DoD contractors only, Administrative or Operational Use, 15 October 2020. Other requests for this document shall be referred to the High Performance Computing Modernization Office, 3909 Halls Ferry Road, Vicksburg, MS 39180.
- Presenter: Mathew Boyer
- Presenter: Wesley Brewer