We are employing machine learning techniques to develop more accurate predictive models that are trained jointly on both simulation and experimental data, comparing more forms of data than is typical, such as images, vector-valued data, and scalars. This research supports the national energy and security missions that rely on predictive models.
Publications and Presentations
Ahn, D. H., et al. 2018. "Flux: Overcoming Scheduling Challenges for Exascale Workflows." WORKS, Dallas, TX, Nov. 2018. LLNL-CONF-756663.
Spears, B. 2018. "Deep Learning: A Guide for Practitioners in the Physical Sciences." Phys. Plasmas. 25. doi: 10.1063/1.5020791. LLNL-JRNL-743970.