Introduction

Welcome to the Uncertainty Quantification Group, in the Department of Aeronautics and Astronautics at MIT. We are part of the Aerospace Computational Design Laboratory, as well as the MIT Center for Computational Science and Engineering and the MIT Statistics and Data Science Center, both within the MIT Schwarzman College of Computing.

Research Overview

Our research focuses on advancing foundational computational methodologies for uncertainty quantification, statistical inference, and machine learning in complex physical systems. Our methodological work is motivated by a wide variety of engineering and environmental applications.

We tackle a broad range of projects, but most involve aspects of a few core questions:

  • How to quantify confidence in computational predictions?
  • How to build or refine models of complex physical processes from indirect and limited observations?
  • What information is needed to drive inference, design, and control?

Featured Publications

O. Zahm, T. Cui, K. Law, A. Spantini and Y. M. Marzouk, Certified dimension reduction in nonlinear Bayesian inverse problems,Mathematics of Computation, in press (2022).
A. Spantini, R. Baptista and Y. M. Marzouk, Coupling techniques for nonlinear ensemble filtering,SIAM Review, in press (2022).
J. Zech and Y. Marzouk, Sparse approximation of triangular transports. Part II: the infinite dimensional case,Constructive Approximation, in press (2022).
J. Jagalur-Mohan and Y. M. Marzouk, Batch greedy maximization of non-submodular functions: guarantees and applications to experimental design,The Journal of Machine Learning Research, 22 (2021), pp. 1–62.
M. Brennan, D. Bigoni, O. Zahm, A. Spantini and Y. M. Marzouk, Greedy inference with structure-exploiting lazy maps,Advances in Neural Information Processing Systems (NeurIPS) oral presentation, (2020).
View all publications