Advancing Materials Characterization Through Physics-Guided Machine Learning
MIE Department Seminar
November 14, 2024
11:00 AM - 12:00 PM America/Chicago
Speaker: Nina Andrejevic, PhD, Argonne National Laboratory
Location: ERF 1043
Abstract: Recent advances in materials characterization techniques and instrumentation are enabling the collection of larger and more sophisticated datasets, creating the possibility for richer and more quantitative understanding of materials structure and properties. However, deriving structural and mechanistic models from data through computational modeling and theory development remains a bottleneck for analysis. Here, I discuss two research directions in which we approach this challenge by leveraging machine learning methods to accelerate the interpretation of materials characterization data. First, I will present our work employing equivariant neural networks to directly predict materials’ vibrational properties from atomic attributes. In addition to enabling efficient, high-throughput screening of materials, such surrogate models can substitute intensive calculations to accelerate the inversion of spectroscopy data to atomic structures. Last, I will share recent progress in developing a data-driven framework to uncover dynamical models from time-resolved coherent X-ray scattering measurements of mesoscale structural dynamics. Using physics-guided neural networks, we enable estimation of long-term behavior well beyond the measurement time and begin to bridge the gap between approximate models and complex data.
Speaker Bio: Nina Andrejevic is a Maria Goeppert Mayer Fellow at Argonne National Laboratory. Her research focuses on developing physics-aware machine learning models for intelligent analysis of materials characterization data. She received her B.S. in engineering physics from Cornell University and her Ph.D. in materials science and engineering from Massachusetts Institute of Technology. Alongside her research, she is also enthusiastic about science communication through teaching and scientific data visualization.
Date posted
Nov 6, 2024
Date updated
Nov 6, 2024