Skip to main content

Applied Math Seminar

Applied Math Seminar

Title:
Mixed Precision Iterative Methods for Linear Inverse Problems
 
Abstract:
Many problems in science and engineering give rise to linear systems of equations that are commonly referred to as large-scale linear discrete ill-posed problems. The matrices that define these problems are typically severely ill-conditioned and may be rank deficient. Because of this, regularization is often needed to stem the effect of perturbations caused by error in the available data. In this talk we consider the solution of the regularized least-squares problem using both mixed precision and Krylov subspace projection techniques. We utilize a filter factor analysis to investigate the regularizing behavior of the proposed iterative methods.
Date:
-
Location:
POT 745

Applied Math Seminar

Title:
Mixed Precision Iterative Methods for Linear Inverse Problems
 
Abstract:
Many problems in science and engineering give rise to linear systems of equations that are commonly referred to as large-scale linear discrete ill-posed problems. The matrices that define these problems are typically severely ill-conditioned and may be rank deficient. Because of this, regularization is often needed to stem the effect of perturbations caused by error in the available data. In this talk we consider the solution of the regularized least-squares problem using both mixed precision and Krylov subspace projection techniques. We utilize a filter factor analysis to investigate the regularizing behavior of the proposed iterative methods.
Date:
-
Location:
POT 745

Applied Math Seminar

Title: Some Modified Matrix Eigenvalue Problems

Abstract: 

This is the title of a well-known numerical linear algebra

survey article by Gene Golub published in 1973. The article

covers a range of matrix eigenvalue problems which require some
manipulations before the standard algorithms may be used.
I am using the same title to consider a new set of modified
matrix eigenvalue problems. This includes constrained and
bi-level optimizations arising from algorithms for fairness in machine
learning, such as spectral clustering with group fairness
and fair principal component analysis. We also consider
eigenvalue optimization via 2D eigenvalue problem with applications
to the calculation of the distance to instability among others,
and stationary values of a quadratic form subject to non-homogeneous
linear constraints for applications such as image segmentation with
constraints. I will discuss how to explore the underlying
structures of these problems to turn them into our familiar
eigenvalue problems and algorithms. 
Date:
-
Location:
POT 745

Applied Math Seminar

Title: Some Modified Matrix Eigenvalue Problems

Abstract: 

This is the title of a well-known numerical linear algebra

survey article by Gene Golub published in 1973. The article

covers a range of matrix eigenvalue problems which require some
manipulations before the standard algorithms may be used.
I am using the same title to consider a new set of modified
matrix eigenvalue problems. This includes constrained and
bi-level optimizations arising from algorithms for fairness in machine
learning, such as spectral clustering with group fairness
and fair principal component analysis. We also consider
eigenvalue optimization via 2D eigenvalue problem with applications
to the calculation of the distance to instability among others,
and stationary values of a quadratic form subject to non-homogeneous
linear constraints for applications such as image segmentation with
constraints. I will discuss how to explore the underlying
structures of these problems to turn them into our familiar
eigenvalue problems and algorithms. 
Date:
-
Location:
POT 745

Applied Math Seminar

Title: Generating Representative Samples: Neural Networks and More

Abstract: Approximating a probability distribution using a discrete set of points is a fundamental task in modern scientific computation, with applications in uncertainty quantification among other things. We discuss recent advances in this area, including the use of Stein discrepancies and various optimization techniques. In particular, we introduce Stein-Message-Passing Monte Carlo (Stein-MPMC), an extension of the original Message-Passing Monte Carlo model and the first machine-learning algorithm for generating low-discrepancy (space-filling) point sets. Additionally, we present a generalized Subset Selection algorithm, a simpler yet highly effective optimization method.

Date:
-
Location:
POT 745

Applied Math Seminar

Title: Generating Representative Samples: Neural Networks and More

Abstract: Approximating a probability distribution using a discrete set of points is a fundamental task in modern scientific computation, with applications in uncertainty quantification among other things. We discuss recent advances in this area, including the use of Stein discrepancies and various optimization techniques. In particular, we introduce Stein-Message-Passing Monte Carlo (Stein-MPMC), an extension of the original Message-Passing Monte Carlo model and the first machine-learning algorithm for generating low-discrepancy (space-filling) point sets. Additionally, we present a generalized Subset Selection algorithm, a simpler yet highly effective optimization method.

Date:
-
Location:
POT 745

Applied Math Seminar

Title: Translating and evaluating single-cell Boolean network interventions in the multiscale setting

Abstract: Intracellular networks process cellular-level information and control cell fate. They can be computationally modeled using Boolean networks, implicit-time causal models of discrete binary events. These networks can be embedded into cell-agents of an agent-based model to drive cellular behavior. To explore this integration, we identify a set of candidate interventions that induce apoptosis in a cell-survival network of a rare leukemia using exhaustive search simulation, stable motif control, and an individual-based mean field approach (IBMFA). Due to algorithm constraints, these interventions are well-suited for cell-level decisions but less so for multicellular agent-based contexts. To address these limitations, we treat the target control solutions as putative therapeutic targets and develop a pipeline to translate them to continuous-time multicellular, agent-based models. We set the discrete-to-continuous transitions between the Boolean network and multicellular model via thresholding and produce simple computational simulations designed to emulate aspects of experimental and translational biology. Our results reveal that interventions performing equivalently in Boolean network simulations diverge in multiscale settings in both population growth and spatial distribution. Further analysis links these differences to internal network dynamics and the intervention’s proximity to output nodes. This proof-of-concept study highlights the importance of accounting for internal dynamics in multicellular simulations and advances understanding of Boolean network control.

Date:
-
Location:
POT 745
Event Series:

Applied Math Seminar

Title: Spatial Dynamics of Vector Borne Diseases

Abstract: Vector-borne diseases affects approximately 1 billion people and accounts for 17% of all infectious diseases. With travel becoming more frequent across the global, it is important to understand the spatial dynamics of vector-borne diseases. Host movement plays a key part on how a disease can be distributed as it enables a pathogen to invade a new environment, and helps the persistence of a disease in locations that would otherwise be isolated. In this talk, we will explore how spatial heterogeneity combines with mobility network structure to influence vector-borne disease dynamics

Date:
-
Location:
POT 745
Tags/Keywords:

Applied Math Seminar

Title: Spatial Dynamics of Vector Borne Diseases

Abstract: Vector-borne diseases affects approximately 1 billion people and accounts for 17% of all infectious diseases. With travel becoming more frequent across the global, it is important to understand the spatial dynamics of vector-borne diseases. Host movement plays a key part on how a disease can be distributed as it enables a pathogen to invade a new environment, and helps the persistence of a disease in locations that would otherwise be isolated. In this talk, we will explore how spatial heterogeneity combines with mobility network structure to influence vector-borne disease dynamics

Date:
-
Location:
POT 745
Tags/Keywords:

Applied Math Seminar

Title: Radiative transport and optical tomography

Abstract: Optical tomography is the process of reconstructing the optical parameters of the inside of an object from measurements taken on the boundary.  This problem is hard if light inside the object is scattered -- if it bounces around a lot and refuses to travel in straight lines. To solve optical tomography problems, we need a mathematical model for light propagation inside a scattering medium. In this talk I'll give a brief introduction to one such model -- the radiative transport model -- and talk a little bit about its behavior and its implications for optical tomography.

Date:
-
Location:
POT 745
Tags/Keywords: