The Scientific Machine Learning (SciML) Lab just published a paper titled, “Convolution Operator Network for Forward and Inverse Problems (FI-Conv): Application to Plasma Turbulence Simulations.” Dr. Xingzhuo Chen, TAMIDS postdoctoral research associate in the SciML Lab led by Dr. Ulisses Braga-Neto, collaborated with the Institute of Fusion Sciences at the University of Texas and the Department of Mathematics at Virginia Tech.
In this paper, Dr. Chen and his collaborators describe FI-Conv, a new artificial intelligence tool that can both predict how complex physical systems will change over time and work backward to figure out the hidden factors controlling them. Their research focused on the application of this tool in plasma turbulence simulations.
This model could offer a powerful alternative to existing physics-based AI models, especially for studying complex systems.
Dr. Chen has other projects in the works as well, including fine-tuning foundation models for supernova research and accelerating radiative transfer simulations using machine learning. Congratulations to Dr. Chen and his team for an exciting contribution to the field of machine learning!
Collaborators
Dr. Anthony Poole – Institute of Fusion Sciences, University of Texas
Dr. David Hatch – Institute of Fusion Sciences, University of Texas
Dr. Ionut-Gabriel Farcas – Department of Mathematics, Virginia Tech