A new research article by Dr. Troy Butler, faculty member in the department, and his co-authors has been published in the International Journal for Uncertainty Quantification. The work also includes contributions from former PhD students Michael Pilosov and Scott Walsh, highlighting the lasting research impact of the department’s graduate alumni.
The article introduces new Optimal Experimental Design (OED) criteria developed specifically for the Data Consistent Inversion (DCI) framework. Unlike traditional OED approaches, which often require solving many computationally expensive inverse problems, the proposed criteria enable efficient experiment selection without performing inverse solves.
By leveraging the geometric structure of data pre-images and utilizing singular value computations of sampled Jacobians, the new methods offer an intuitive and computationally efficient pathway for reducing uncertainty in model predictions. The authors also present both simultaneous and sequential (greedy) algorithms.
This publication reflects the department’s strength in faculty-led research, graduate mentorship, and methodological innovation in uncertainty quantification and computational science. We congratulate Dr. Butler and the full author team on this achievement.