Search Research Symposium Abstracts

Flexible Krylov Methods for Group Sparsity Regularization

In this talk, I will discuss the concept of group sparsity regularization, which involves adding structured sparsity to the regularization process by assigning variables to predefined groups, that can be overlapping or non-overlapping. Next, I will present novel hybrid projection methods based on flexible Krylov subspaces. The approach begins by transforming [Read More...]

Presenter: Malena Sabaté Landman, Emory University
Authors: Julianne Chung
Symposium Year: 2023
Session: Computational Inverse Problems and Uncertainty Quantification [Organized by Julianne Chung, Rosemary Renaut, and Malena Sabate-Landman]
Presentation Time: September 30, 2023; 2:00 pm

Phase and absorption contrast imaging using intensity measurements

We consider imaging absorbing as well as non-absorbing objects using intensity only measurements. Objects with high absorption contrast can be imaged effectively using multiple illuminations and/or masks as in ghost imaging. On the other hand, transparent objects with low absorption contrast are more challenging to be imaged when only intensities are [Read More...]

Presenter: Chrysoula Tsogka, University of California Merced
Authors: M. Moscoso, A. Novikov, G. Papanicolaou and C. Tsogka
Symposium Year: 2023
Session: Computational Inverse Problems and Uncertainty Quantification [Organized by Julianne Chung, Rosemary Renaut, and Malena Sabate-Landman]
Presentation Time: September 30, 2023; 2:25 pm

RECYCLING MMGKS FOR LARGE-SCALE DYNAMIC AND STREAMING DATA

The ubiquity of inverse problems in many fields of science is associated with emerging challenges on obtaining relevant solutions to large-scale and data-intensive inverse problems such as ill-posedness, large dimensionality of the parameters, and the complexity of the model constraints. Edge-preserving constraint has received considerable attention due to [Read More...]

Presenter: Mirjeta Pasha, Tufts University
Authors: Eric de Sturler and Misha Kilmer
Symposium Year: 2023
Session: Computational Inverse Problems and Uncertainty Quantification [Organized by Julianne Chung, Rosemary Renaut, and Malena Sabate-Landman]
Presentation Time: September 30, 2023; 2:50 pm

Statistical Learning Theory of Deep Neural Networks for Data with Low-Dimensional Structures

In the past decade, deep learning has made astonishing breakthroughs in various real-world applications. It is a common belief that deep neural networks are good at learning various geometric structures hidden in data sets, such as rich local regularities, global symmetries, or repetitive patterns. One of the central interests in deep learning theory is to [Read More...]

Presenter: Wenjing Liao, Georgia Institute of Technology
Symposium Year: 2023
Session: Computational Inverse Problems and Uncertainty Quantification [Organized by Julianne Chung, Rosemary Renaut, and Malena Sabate-Landman]
Presentation Time: September 30, 2023; 3:15 pm

Hierarchical Bayesian Filtering for Time-Varying Parameter Estimation in Dynamical Systems

Estimating and quantifying uncertainty in unknown system parameters from limited, noisy data remains a challenging inverse problem. In addition to constant (static) parameters, a variety of systems include unobservable parameters that vary with time but have unknown evolution models. In this talk, we present a particle filtering method that utilizes a [Read More...]

Presenter: Andrea Arnold, Worcester Polytechnic Institute
Symposium Year: 2023
Session: Computational Inverse Problems and Uncertainty Quantification [Organized by Julianne Chung, Rosemary Renaut, and Malena Sabate-Landman]
Presentation Time: October 1, 2023; 2:00 pm

Hybrid Iterative Solver for Inverse Problems

Inverse problems arise in a variety of applications: machine learning, image processing, finance, mathematical biology, and more. Solution schemes are formulated by applying algorithms that incorporate regularization techniques and/or statistical approaches. In most cases these solution schemes involve the need to solve large-scale ill-conditioned linear [Read More...]

Presenter: Ariana Brown, Emory University
Authors: Ariana Brown, James Nagy, Malena Sabate Landman
Symposium Year: 2023
Session: Computational Inverse Problems and Uncertainty Quantification [Organized by Julianne Chung, Rosemary Renaut, and Malena Sabate-Landman]
Presentation Time: October 1, 2023; 2:25 pm

Training Data for Geophysical Inversion

Geophysical inversion is challenging because realistic models of the Earth are nonlinear and their corresponding inverse problems are ill-posed due to non-existent or nonunique solutions that are sensitive to small changes in the data. There is now a significant body of work in geophysics that uses neural networks to represent an inverse operator. It can be [Read More...]

Presenter: Jodi Mead, Boise State University
Symposium Year: 2023
Session: Computational Inverse Problems and Uncertainty Quantification [Organized by Julianne Chung, Rosemary Renaut, and Malena Sabate-Landman]
Presentation Time: October 1, 2023; 2:50 pm

Improving inversion of geophysical potential field data sets using variable projection

Geophysics data from subsurface geometries give rise to magnetic and gravity potential fields that are measured only above, or at the surface, of the region to be recovered. The aim in the inversion is to reconstruct the subsurface structures in terms of magnetic susceptibility or density, respectively. The inversion problem is highly under determined; there [Read More...]

Presenter: Rosemary Renaut, Arizona State University
Authors: Rosemary Renaut, Matthias Chung and Saeed Vatankhah
Symposium Year: 2023
Session: Computational Inverse Problems and Uncertainty Quantification [Organized by Julianne Chung, Rosemary Renaut, and Malena Sabate-Landman]
Presentation Time: October 1, 2023; 3:15 pm