Artificial Intelligence and Machine Learning (AI/ML)

The Artificial Intelligence/Machine Learning (AI/ML) research activities at the School of Physics involve designing machine learning and reasoning algorithms to accelerate physics research, as well as using techniques from physics to develop reliable AI algorithms. Faculty in the School design AI/ML algorithms applicable to a variety of settings, including astrophysics, cosmology, dynamical systems, hydrodynamics, fundamental physics, and neuroscience. These algorithms incorporate first-principles physics concepts, uncertainty quantification, and interpretability tools, ingest data to discover new dynamical equations, or use Large Language Models (LLMs), leveraging AI supercomputers at scale.

AI research in SoP benefits from interdisciplinary connections through the AI4Science Center (ai4science.ai.gatech.edu), the Institute for Data Engineering and Science (IDEaS; https://research.gatech.edu/data), Tech AI (https://ai.gatech.edu), and the ML Center (https://ml.gatech.edu) at Georgia Tech.

 

 
 
 

Faculty Members

Flavio Fenton

Professor

Research Interests: Theory and simulation of complex cardiometrics.

Aishik Ghosh

Assistant Professor

Research Interests: Neuro-symbolic AI, high-dimensional statistics, uncertainty quantification, fast simulation, fast inference & experiment design in the context of fundamental physics and astrophysics

Roman Grigoriev

Professor

Research Interests: Model discovery, fluid dynamics, nonlinear systems

Feryal Ozel

Professor and Chair

Research Interests: Interferometric imaging, multi-modal data, black holes

Dimitrios Psaltis

Professor and Director of the AI4Science Center

Research Interests: AI methods in astrophysics, HPC, general relativity and black holes

Audrey Sederberg

Assistant Professor

Research Interests: Theoretical neuroscience, artificial and biological neural network models

Ignacio Taboada

Professor

Research Interests: Neutrino Astrophysics with IceCube and P-ONE. AI methods in neutrino event reconstruction; AI methods on event selection for neutrino telescopes.

John Wise

Professor

Research Interests: HPC cosmological simulations, radiation transport, black holes, AI methods in simulations – physics emulators, physics-informed NNs, multi-modal models