Share
Zhe Bai
Zhe Bai is a computational science researcher in the Computing Sciences Area at Lawrence Berkeley National Laboratory. Her research interests lie in the area of sparse sampling and model order reduction, including compressed sensing, machine learning and large-scale computation and simulation. Cultivated interdisciplinary research and collaborations spanning the fields of engineering and applied mathematics, her work involves data-driven modeling that leverages advanced data science techniques to understand, estimate and control high-dimensional physical systems.
Education
- Ph.D., Mechanical Engineering, University of Washington, 2018.
- M.S., Applied Mathematics, University of Washington, 2017.
- M.S., Mechanical & Aerospace Engineering, Syracuse University, 2013.
- B.S., Thermal Energy & Power Engineering, Harbin Institute of Technology, 2011.
Previous Appointments
- Research Assistant, University of Washington, Seattle, WA.
- Research Intern, Sandia National Laboratories, Livermore, CA.
- Research Assistant, Syracuse University, Syracuse, NY.
Selected Publications
Journal Articles
Zhe Bai, Xishuo Wei, William Tang, Leonid Oliker, Zhihong Lin, Samuel Williams, "Transfer Learning Nonlinear Plasma Dynamic Transitions in Low Dimensional Embeddings via Deep Neural Networks", Machine Learning: Science and Technology, April 8, 2025, doi: 10.1088/2632-2153/adca83
Mustafa Mutiur Rahman, Zhe Bai, Jacob Robert King, Carl R. Sovinec, Xishuo Wei, Samuel Williams, Yang Liu, "Sparsified time-dependent Fourier neural operators for fusion simulations", Phys. Plasmas, December 4, 2024, 31:12, doi: 10.1063/5.0232503
Á Sánchez-Villar, Z Bai, N Bertelli, EW Bethel, J Hillairet, T Perciano, S Shiraiwa, GM Wallace, JC Wright, "Real-time capable modeling of ICRF heating on NSTX and WEST via machine learning approaches", Nuclear Fusion, August 12, 2024, 64:9, doi: 10.1088/1741-4326/ad645d
Zhe Bai, Abdelilah Essiari, Talita Perciano, Kristofer E Bouchard, "AutoCT: Automated CT registration, segmentation, and quantification", Software X, February 28, 2024, 26, doi: 10.1016/j.softx.2024.101673
Gregory Wallace, Zhe Bai, Robbie Sadre, Talita Perciano, Nicola Bertelli, Syun'ichi Shiraiwa, Wes Bethel, John Wright, "Towards fast and accurate predictions of radio frequency power deposition and current profile via data-driven modelling: applications to lower hybrid current drive", Journal of Plasma Physics, August 18, 2022, 88:4, doi: 10.1017/S0022377822000708
M. Avaylon, R. Sadre, Z. Bai, T. Perciano, "Adaptable Deep Learning and Probabilistic Graphical Model System for Semantic Segmentation", Advances in Artificial Intelligence and Machine Learnin, March 31, 2022, 2:288--302, doi: 10.54364/AAIML.2022.1119
Zhe Bai, Liqian Peng, "Non-intrusive nonlinear model reduction via machine learning approximations to low-dimensional operators", Advanced Modeling and Simulation in Engineering Sciences, 2021, 8:28, doi: 10.1186/s40323-021-00213-5
Zhe Bai, N. Benjamin Erichson, Muralikrishnan Gopalakrishnan Meena, Kunihiko Taira, Steven L. Brunton, "Randomized methods to characterize large-scale vortical flow networks", PLOS ONE Journal, 2019, doi: 10.1371/journal.pone.0225265
Zhe Bai, Eurika Kaiser, Joshua L. Proctor, J. Nathan Kutz, Steven L. Brunton, "Dynamic mode decomposition for compressive system identification", AIAA Journal, 2019, 58:2, doi: 10.2514/1.J057870
Zhe Bai, Thakshila Wimalajeewa, Zachary Berger, Guannan Wang, Mark Glauser, Pramod K Varshney, "Low-dimensional approach for reconstruction of airfoil data via compressive sensing", AIAA Journal, April 2015, 53:4, doi: 10.2514/1.J053287
Conference Papers
Damian Rouson, Zhe Bai, Dan Bonachea, Kareem Ergawy, Ethan Gutmann, Michael Klemm, Katherine Rasmussen, Brad Richardson, Sameer Shende, David Torres, Yunhao Zhang, "Automatically parallelizing batch inference on deep neural networks using Fiats and Fortran 2023 `do concurrent`", Fifth International Workshop on Computational Aspects of Deep Learning (CADL), June 2025, doi: 10.25344/S4VG6T
This paper introduces novel programming strategies that leverage features of the Fortran 2023 standard of the International Standards Organization (ISO) to automatically parallelize computations on deep neural networks. The paper focuses on the interplay of object-oriented, parallel, and functional programming paradigms in the Fiats deep learning library. We demonstrate how several infrequently used language features play a role in enabling efficient, parallel execution. Specifically, the ability to explicitly declare that a procedure is pure facilitates inference in the context of the language’s loop-parallelism construct `do concurrent`. Also, explicitly prohibiting the overriding of a parent type’s type-bound procedures eliminates the need for dynamic dispatch in performance-critical code. Finally, this paper uses batch inference calculations on a neural network surrogate for atmospheric aerosol dynamics to demonstrate that LLVM Flang compiler’s automatic parallelization of `do concurrent` achieves roughly the same performance and scalability as achieved by OpenMP compiler directives. We also demonstrate that double-precision inference costs 37–72% longer runtime than default-real precision with most values in the range 57-60%.
