Dr. Braden C. Soper is a data scientist and group leader at Lawrence Livermore National Laboratory (LLNL), where he oversees the Informatics Group within the Center for Applied Scientific Computing. With a technical background in applied mathematics and statistics, Dr. Soper's work focuses on developing scalable, interpretable machine learning and statistical models for real-world data, especially in domains critical to public health and national security.

Dr. Soper earned his Ph.D. in Applied Mathematics and Statistics from the University of California, Santa Cruz, and his B.S. in Mathematics from UCLA, where he graduated summa cum laude. During his graduate years, he worked as a researcher at both LLNL and NOAA’s National Marine Fisheries Service, honing his skills in Bayesian modeling, stochastic processes, dynamical systems, and statistical inference.

Since joining LLNL full-time in 2016, Dr. Soper has led interdisciplinary teams addressing problems across biomedical analytics, cybersecurity, nuclear deterrence, and autonomous systems. His contributions include:

• Healthcare analytics: Development of probabilistic machine-learning models for dynamic risk stratification tools using electronic health records (EHR) from institutions such as Kaiser Permanente, University of Toledo, and the Veterans Affairs health system.
• Cancer screening modeling: Leadership in collaborations with the Cancer Registry of Norway and other partners to build Bayesian models for predictive oncology, improving screening strategies and survival forecasting across HPV-related cancers.
• COVID-19 modeling: Creation of dynamic models that revealed disease state–dependent risk factors for hospitalized patients, with publications in JAMIA and Scientific Reports.
• National security modeling: Game-theoretic and reinforcement learning approaches for analyzing strategic stability in the context of nuclear deterrence, leading to briefings for U.S. government stakeholders.
• Sensor networks and decision theory: Algorithms for decentralized Bayesian learning and cooperative estimation, advancing capabilities for multi-agent reinforcement learning in autonomous platforms.

 

Ph.D. Applied Mathematics and Statistics, University of California, Santa Cruz, CA

B.S. Mathematics, University of California, Los Angeles, CA

Selected Publications

Soper, B., Miller, C., & Merl, D. (2024). Linearly solvable general-sum Markov games. Proceedings of the 60th Annual Allerton Conference on Communication, Control, and Computing.

Bahney, B., & Soper, B. (2024). The soft side of superiority: Damage limitation and qualitative advantage in future U.S.–China crises. In B. Roberts (Ed.), Counterforce in Contemporary U.S. Nuclear Strategy (pp. 234–244). Center for Global Security Research, Lawrence Livermore National Laboratory.

Soper, B., Ray, P., Chen, H., Cadena, J., & Goldhahn, R. (2022). Bayesian multiagent active sensing and localization via decentralized posterior sampling. Proceedings of the 56th Annual Asilomar Conference on Signals, Systems, and Computers.

Soper, B., Cadena, J., Nguyen, S., Chan, K. H. R., Kiszka, P., Womack, L., Work, M., Duggan, J., Haller, S. T., Hanrahan, J. A., Kennedy, D. J., Mukundan, D., & Ray, P. (2022). Dynamic modeling of hospitalized COVID-19 patients reveals disease state–dependent risk factors. Journal of the American Medical Informatics Association, 29(5), 864–872.

Meng, R., Soper, B., Lee, H. K. H., Nygard, J., & Nygard, M. (2022). Hierarchical continuous-time inhomogeneous hidden Markov model for cancer screening data. Statistical Methods in Medical Research, 31(8), 1585–1600.

Meng, R., Soper, B., Lee, H. K. H., Liu, V. X., Greene, J. D., & Ray, P. (2021). Nonstationary multivariate Gaussian processes for electronic health records. Journal of Biomedical Informatics, 117, 103698.

Nguyen, S., Chan, K. H. R., Cadena, J., Soper, B., Kiszka, P., Womack, L., Work, M., Duggan, J., Haller, S. T., Hanrahan, J. A., Kennedy, D. J., Mukundan, D., & Ray, P. (2021). Budget-constrained machine learning for early prediction of adverse outcomes for COVID-19 patients. Scientific Reports, 11, 20581.

Soper, B., Nygard, M., Abdulla, G., Meng, R., & Nygard, J. (2020). A hidden Markov model for population-level cancer screening data. Statistics in Medicine, 39(25), 3569–3590.

Goncalves, A., Soper, B., Ray, P., Widemann, D., Nygard, M., Nygard, J., & Sales, A. P. (2020). Improving five-year survival predictions via multitask learning across HPV-related cancers. PLOS ONE, 15(11), e0241225.

Goncalves, A., Ray, P., Soper, B., Stevens, J., Coyle, L., & Sales, A. P. (2020). Generation and evaluation of synthetic patient data. BMC Medical Research Methodology, 20, 108.

Goncalves, A., Ray, P., Soper, B., Widemann, D., Nygard, M., Nygard, J., & Sales, A. P. (2019). Bayesian multitask learning regression for heterogeneous patient cohorts. Journal of Biomedical Informatics: X, 4, 100063.

Soper, B. (2019). A cyber-nuclear deterrence game. Proceedings of the 57th Annual Allerton Conference on Communication, Control, and Computing, 116–123.

Soper, B., & Musacchio, J. (2015). A non-zero-sum, sequential detection game. Proceedings of the 53rd Annual Allerton Conference on Communication, Control, and Computing, 1136–1143.

Soper, B., & Musacchio, J. (2014). A botnet detection game. Proceedings of the 52nd Annual Allerton Conference on Communication, Control, and Computing, 577–584.