Brian Kiedrowski

Dr.Kiedrowski’s research focuses on the development of methods related to understanding the interactions and transport of particle radiation (neutrons, photons, and charged particles) within matter. To solve these problems, he employs both traditional numerical linear algebra methods and Monte Carlo simulations and apply data-driven approaches for the purpose of acceleration and uncertainty quantification. Specific projects include using Bayesian optimization methods to predictive and uncertainty quantified optical models for nuclear fission, applying techniques from computer vision with low-fidelity simulations to categorize transport problems for variance reduction technique selection for high-fidelity Monte Carlo calculations, using tensor network decomposition methods to accelerate transport simulations in multiscale or hierarchical problems, and using multi-fidelity transport methods (diffusion, SPN, and transport).