Andrey A. Popov

Assistant Professor (U.S. Citizen)

Teaching

• Spring 2026: ICS 141 Discrete Math for CS I
• Spring 2026: ICS 241 Discrete Math for CS II
• Fall 2025: ICS 141 Discrete Math for CS I
• Spring 2025: ICS 691B Intro to Data Fusion
• Fall 2024: ICS 141 Discrete Math for CS I

Research

Multifidelity inference

How can we make use of computer models that are untrustworthy, like those created by data-driven methods such as neural
networks? Multifidelity and model forest methods build a hierarchy of models where the principal, top-level, model is
based on theory and expert understanding, meaning that it is trustworthy. The subsequent lower level models are all less
trustworthy than the principal model, meaning that these so-called ancillary models can for-go theory and instead be based
on data. Models such as neural-networks, and reduced-order models can be very cheap for the predictive power that they
provide, but are much less interpretable and much less trustworthy.

In the figure above, the bifidelity ensemble filtering architecture is described. Prior information is forecasted through
two different models. The first being a trustworthy "supervisor" model, with the second being an untrustworthy "subordinate"
model. That information is combined into one source of information, begetting the total prior information. Observations from
sensors are then filtered together with this total prior information to beget our posterior information. Samples from this
total posterior information are taken (through either deterministic of stochastic means) and the cycle repeats.

In the figure above, the bifidelity ensemble Kalman filter architecture with reduced order modeling is described. The top-level
full order model has two samples which are transformed into the space of the reduced order model, creating what is known as a
control variate. The information from the full order model and the 12 samples of the reduced order model is coupled through the
use of this control variate, and a Kalman gain is constructed. This Kalman gain is used to update the full order model and
reduced order model ensembles. The ensemble members are propagated and the cycle repeats.

In the figure above, the model forest ensemble Kalman filter architecture is described. This generalizes the above bifidelity
ensemble Kalman filter by having the possibility of multiple top-level full order models, and multiple layers and branches of
lower-level subordinate reduced order models.

Publications

• Popov, Andrey A., Changhong Mou, Adrian Sandu, and Traian Iliescu. "A multifidelity ensemble Kalman filter with reduced order control variates." SIAM Journal on Scientific Computing 43, no. 2 (2021): A1134-A1162.
• Popov, Andrey A., and Adrian Sandu. "Multifidelity data assimilation for physical systems." In Data Assimilation for Atmospheric, Oceanic and Hydrologic Applications (Vol. IV), pp. 43-67. Cham: Springer International Publishing, 2022.
• Popov, Andrey A., and Adrian Sandu. "The Model Forest Ensemble Kalman Filter." arXiv preprint arXiv:2210.11971 (2022).
• Popov, Andrey A., and Adrian Sandu. "Multifidelity ensemble Kalman filtering using surrogate models defined by theory-guided autoencoders." Data-driven modeling and optimization in fluid dynamics: From physics-based to machine learning approaches 16648714 (2023): 41.

Aerospace Applications

Publications

• "Giraldo-Grueso, Felipe; Popov, Andrey A; Zanetti, Renato; ",Precision Mars Entry Navigation with Atmospheric Density Adaptation via Neural Networks,Journal of Aerospace Information Systems,21,12,982-995,2024,American Institute of Aeronautics and Astronautics
• "Durant, Dalton; Popov, Andrey A; Zanetti, Renato; ",What are You Weighting For? Improved Weights for Gaussian Mixture Filtering With Application to Cislunar Orbit Determination,arXiv preprint arXiv:2405.11081,2024,
• "Giraldo-Grueso, Felipe; Popov, Andrey A; Zanetti, Renato; ",Gaussian Mixture-Based Point Mass Filtering With Applications to Terrain-Relative Navigation,IEEE Transactions on Aerospace and Electronic Systems,2025,IEEE
• "Popov, Andrey A; Zanetti, Renato; ",Deterministic Optimal Transport-based Gaussian Mixture Particle Filtering for Verifiable Applications,arXiv preprint arXiv:2501.17302,2025,
• "Giraldo-Grueso, Felipe; Popov, Andrey A; Hanebeck, Uwe D; Zanetti, Renato; ",Optimal Sampling for Point Mass Filtering with Applications to Cislunar Orbit Determination,The Journal of the Astronautical Sciences,72,5,44,2025,Springer US New York
• "Durant, Dalton; Popov, Andrey A; DeMars, Kyle J; Zanetti, Renato; ",A Framework for Batch Processing Tracklets in the GM-PHD Filter,Advances in Space Research,2025,Elsevier
• "Giraldo-Grueso, Felipe; Popov, Andrey A; Rea, Jeremy R; Zanetti, Renato; ",Adaptive Atmospheric Entry Guidance via Neural Networks,"Journal of Guidance, Control, and Dynamics",2025,American Institute of Aeronautics and Astronautics

Prospective Students

Software Projects

Media