Seminar Series

About the Series

The SIAM PNW Section Seminar is a quarterly seminar presented by a distinguished researcher, invited by the officers of the section. Suggestions for future speakers are welcomed.

Due to the large geographic extent of our section, the seminar series consists of online talks only, currently one per academic quarter.  If you would like to virtually attend these talks, please contact one of the section officers or your local SIAM PNWS contact.

A YouTube Playlist with recordings of some of the previous seminars is available here: https://www.youtube.com/playlist?list=PLAENrdU3Q0mL3SmGN9MGrq8kTI6wRtQ8f . Recordings are also embedded in the page below.

Upcoming Seminars

Date and time: Tuesday 4:00 pm Pacific, April 23, 2024
https://sfu.zoom.us/j/87325983335?pwd=b0mG16JEkDAjKkbxPP1YFsrimstYuz.1 


Speaker: Lisa Fauci (Tulane)

Title: Unraveling hydrodynamic performance at the microscale:  perspectives on modeling

Abstract: Choanoflagellates,  important protists in aquatic ecosystems, exhibit a distinctive cellular structure consisting of a cell body, a flagellum, and a collar of microvilli enveloping the flagellum. Hydrodynamics governs the feeding performance of these organisms, that can be unicellular or form colonies.   They use their flagella to swim and to create a water current to transport bacterial prey to their food-capturing collar.  Remarkably, certain species, such as C. Flexa, can dynamically invert the morphology of their colony by contracting or relaxing the actin rings of the collars of individual cells comprising the colony.  We will share our recent work that  coordinates laboratory experiments and simulations to investigate the swimming and feeding performance of individual choanoflagellates carrying bacterial prey, as well the hydrodynamic performance of colonies of hundreds of these cells.   We will discuss modeling choices, computational frameworks, and successes and challenges. 


Previous Seminars

Spring 2023


Speaker: Per-Gunnar Martinsson (UT Austin)

Title: Randomized algorithms for linear algebraic computations

Abstract: The talk will describe how randomized algorithms can effectively, accurately, and reliably solve linear algebraic problems that are omnipresent in scientific computing and in data analysis. We will focus on techniques for low rank approximation, since these methods are particularly simple and powerful. The talk will also briefly survey a number of other randomized algorithms for tasks such as solving linear systems, estimating matrix norms, and approximating global operators from scientific computing and machine learning.




Winter 2022


Speaker: Prof. Bei Wang (University of Utah)

Title: Hypergraphs: Topological Simplification and Co-Optimal Transport

Abstract:

Hypergraphs capture multi-way relationships in data, and they have consequently seen a number of applications in higher-order network analysis, computer vision, geometry processing, and machine learning. First, we study hypergraph visualization via its topological simplification. We put vertex simplification and hyperedge simplification in a unifying framework using tools from topological data analysis. In simplifying a hypergraph, we allow vertices to be combined if they belong to almost the same set of hyperedges, and hyperedges to be merged if they share almost the same set of vertices. Second, we develop the theoretical foundations in studying the space of hypergraphs using ingredients from optimal transport. By enriching a hypergraph with probability measures on its nodes and hyperedges, as well as relational information capturing local and global structure, we obtain a general and robust framework for studying the collection of all hypergraphs. We first introduce a hypergraph distance based on the co-optimal transport framework of Redko et al. and study its theoretical properties. We then formalize common methods for transforming a hypergraph into a graph as maps from the space of hypergraphs to the space of graphs and study their functorial properties and Lipschitz bounds. Finally, we demonstrate the versatility of our Hypergraph Co-Optimal Transport (HyperCOT) framework through various examples. This talk is based on joint works with Youjia Zhou, Archit Rathore, Emilie Purvine, Samir Chowdhury, Tom Needham, and Ethan Semrad. See https://arxiv.org/abs/2104.11214 and https://arxiv.org/abs/2112.03904.



Fall 2021


Speaker: Margot Gerritsen (Prof., Department of Energy Resources Engineering, Stanford University)

 

Time: Tuesday, Nov 16, 2021 04:00 PM Pacific Time

Title: Nou breekt mijn klomp: 40 years of adventures in STEM

Translation : “What the heck just happened : 40 years of adventures in STEM"

 

Abstract:  Dr. Gerritsen will reflect on her 40 years in STEM fields, modeling in geophysics, coastal ocean dynamics, yacht design, wildfire prediction and mitigation. She will also talk about her work as Co-Director of Women in Data Science, and data science in general.  


Spring 2021

Speaker: Ted Lystig, Medtronic

Time: Tuesday, April 6, 4 pm PT

Talk Title: Ethical allocation of ventilators at the start of the COVID-19 pandemic

Abstract: In early 2020 as the COVID-19 pandemic was spreading across the US, a notable concern was how manufacturers such as Medtronic should allocate their newly produced ventilators to different parts of the country.  Configuration and shipment decisions for these life-saving products needed to be made weeks ahead of time, and we did not yet have great understanding on how disease prevalence was likely to change over time.  I this presentation I will discuss the evolution of multiple competing models to describe the spread of COVID-19, and talk about how we leveraged the output from these models to develop an ethical approach to the allocation of ventilators.  While some technical details will be provided, the talk as a whole will be broadly accessible to college students without extensive specialized training. 

Winter 2020

Speaker: Jacob Bedrossian, University of Maryland

Time: Thursday, February 20, 4pm PST

Talk Title: The power spectrum of passive scalar turbulence in the Batchelor regime

Abstract: In 1959, Batchelor predicted that passive scalars advected in fluids at finite Reynolds number with small diffusivity κ should display a |k|−1 power spectrum over a small-scale inertial range in a statistically stationary experiment. This prediction has been experimentally and numerically tested extensively in the physics and engineering literature and is a core prediction of passive scalar turbulence. Together with Alex Blumenthal and Sam Punshon-Smith, we have provided the first mathematically rigorous proof of this prediction for a scalar field evolving by advection-diffusion in a fluid governed by the 2D Navier-Stokes equations and 3D hyperviscous Navier-Stokes equations in a periodic box subjected to stochastic forcing at arbitrary Reynolds number. These results are proved over the course of four papers by studying the Lagrangian flow map using infinite dimensional extensions of ideas from random dynamical systems. We prove that the Lagrangian flow has a positive Lyapunov exponent (Lagrangian chaos) and show how this can be upgraded to almost sure exponential (universal) mixing of passive scalars at zero diffusivity and further to uniform-in-diffusivity mixing.

Spring 2019

Speaker: Professor David Gleich, Purdue University

Time: Thursday, April 25, 4pm PST

Title: Higher-order clustering of complex networks.

Spectral clustering is a well-known way to partition a graph or network into clusters or communities with provable guarantees on the quality of the clusters. This guarantee is known as the Cheeger inequality and it holds for undirected graphs. We'll discuss a new generalization of the Cheeger inequality to higher-order structures in networks including network motifs. This is easy to implement and seamlessly generalizes spectral clustering to directed, signed, and many other types of complex networks. In particular, our generalization allow us to re-use the large history of existing ideas in spectral clustering including local methods, overlapping methods, and relationships with kernel k-means. We will illustrate the types of clusters or communities found by our new method in biological, neuroscience, ecological, transportation, and social networks.

Papers and Code: https://arxiv.org/pdf/1612.08447.pdf (in Science, 2016), https://arxiv.org/abs/1704.05982 (at KDD2017), https://github.com/arbenson/higher-order-organization-julia

Winter 2019

Speaker: Professor Caroline Uhler, MIT

Time: Thursday, January 31st, 3pm PST

Title: From Causal Inference to Gene Regulation

A recent break-through in genomics makes it possible to perform perturbation experiments at a very large scale.

The availability of such data motivates the development of a causal inference framework that is based on

observational and interventional data. We first characterize the causal relationships that are identifiable from

interventional data. In particular, we show that imperfect interventions, which only modify (i.e., without

necessarily eliminating) the dependencies between targeted variables and their causes, provide the same causal

information as perfect interventions, despite being less invasive. Second, we present the first provably

consistent algorithm for learning a causal network from a mix of observational and interventional data. This

requires us to develop new results in geometric combinatorics. In particular, we introduce DAG associahedra, a

family of polytopes that extend the prominent graph associahedra to the directed setting. We end by discussing

applications of this causal inference framework to the estimation of gene regulatory networks.

Fall 2018

Speaker: Professor Luminita Vese, UCLA

Time: Thursday, October 4, 4pm PST

Title: Variational Methods in Image Processing

Abstract: Many inverse problems arising in image processing can be solved by variational methods. A functional is minimized, composed of a data fidelity term and a regularizing term. The regularizing term imposes a-priori constraints on the solution and makes the problem well-posed. The unknown is found in the appropriate space of functions that best characterizes the desired properties of the solution. The Euler-Lagrange equation associated with the minimization is obtained and then solved using numerical methods. In this talk I will present several inverse problems arising in imaging applications and their solutions in a variational approach. Examples include image restoration, segmentation, decomposition, and medical applications such as image reconstruction in computer tomography. Theoretical and experimental results will be presented.

Winter 2018

Speaker: Professor Leah Keshet, University of British Columbia

Time: Thursday, January 11, 3pm PST

Title: Mathematical models for cell shape

Abstract: In this talk I will describe work carried out in my group on the way that cell shape and motility is regulated. We have developed models for regulatory proteins (GTPases) that control the polarity and shape (contracted or expanded) of mammalian cells such as white blood cells. The models lead to interesting mathematics, as well as insight into behaviour observed in experimental cell biology. I will review both partial and ordinary differential equation versions of our models, and illustrate some new methods devised to analyse these. I will conclude with examples of ongoing collaboration with cell biologists.

Fall 2017

Speaker: Professor Jodi Mead, Boise State University

Time: Thursday, December 7, 3pm PST

Title: Singular value analysis of Joint Inversion

Abstract: The degree of ill-posedness of the linear operator equation Ax = b can be measured by the decay rate of the singular values of A. If the inverse problem is nonlinear and A is the corresponding Jacobian matrix, the singular values measure the local ill-posedness of the problem. We consider the case where the operator A is compact and maps infinite dimensional Hilbert spaces. The degree of ill-posedness can be measured by μ, where the singular values σ_n(A) ≈ n^(−μ). The problem becomes more difficult to solve numerically with increasing μ.

When the singular values quickly decay to zero, it is common to truncate them or introduce a regularization term. However, if the problem is severely ill-posed, a large amount of regularization may be required to solve the problem, and this can introduce significant bias error in the solution estimate. Regularization can be viewed as adding information to the ill-posed problem, and hence we consider the regularized inverse problem as simultaneous joint inversion. Simultaneous joint inversion has recently become a common method to incorporate multiple types of data and physics in a single inversion.

We extend discrete techniques of stacking matrices in joint inversion, to combining Green’s function solutions of multiple differential equations representing different types of data. The singular values of the joint operators indicate the effectiveness of combining multiple types of physics. This knowledge provides mathematical justification for joint inversion, and can be determined before the complicated machinery of discretizing and solving the problem is implemented. We will give an example of two differential equations with known Green’s function solutions. The decay rate of the singular values of the individual operators are compared to the singular values of the joint operator, and the extent to which the ill-posedness was resolved is quantified.

This is joint work with James Ford. 

Spring 2017

Speaker: Professor Ralph Showalter, Oregon State University

Time: Thursday, May 4, 3pm PST

Title: A pseudo-parabolic PDE for compaction of a sedimentary basin 

Abstract: The porosity of a compacting sedimentary basin satisfies a nonlinear pseudo-parabolic partial differential equation.  This equation is distinguished from the classical porous medium equation by a third order term, a degenerate elliptic operator in spatial variables acting on the time derivative of the solution. We describe the derivation of the model equation, review classical results for pseudo-parabolic equations and their relation to parabolic equations, present new existence and regularity results for this nonlinear PDE, and show they are consistent with expected behavior of solutions in this context.

Slides: 17_SIAM_PNW.pdf

Winter 2017

Speaker: Professor Nilima Nigam, Simon Fraser University

Time: Thursday, January 26, 3pm PST

Title: A modification of Schiffer's conjecture, and a proof via finite elements

Abstract: Approximations via conforming and non-conforming finite elements can be used to construct validated and computable bounds on eigenvalues for the Dirichlet Laplacian in certain domains. If these are to be used as part of a proof, care must be taken to ensure each step of the computation is validated and verifiable. In this talk we present a long-standing conjecture in spectral geometry, and its resolution using validated finite element computations.  

Schiffer’s conjecture states that if a bounded domain Ω in R^n has any nontrivial Neumann eigenfunction which is a constant on the boundary, then Ω must be a ball. This conjecture is open. A modification of Schiffer’s conjecture is: for regular polygons of at least 5 sides, we can demonstrate the existence of a Neumann eigenfunction which is does not change sign on the boundary. In this talk, we provide a recent proof using finite element calculations for the regular pentagon. The strategy involves iteratively bounding eigenvalues for a sequence of polygonal subdomains of the triangle with mixed Dirichlet and Neumann boundary conditions. We use a learning algorithm to find and optimize this sequence of subdomains, and use non-conforming linear FEM to compute validated lower bounds for the lowest eigenvalue in each of these domains. The linear algebra is performed within interval arithmetic. This is joint work with Bartlomiej Siudeja and Ben Green, at U. Oregon.

Slides: SIAMTalk.pdf

Fall 2016

Speaker: Professor Randy LeVeque, Applied Mathematics, University of Washington

Time: Thursday, October 20, 3pm PST

Title: New tools for tsunami warning and probabilistic hazard assessment 

Abstract: As events of the past decade have tragically demonstrated, tsunamis pose a major risk to coastal populations around the world -- including the Pacific Northwest where the Cascadia Subduction Zone unleashes Magnitude 9 earthquakes every few hundred years.  Numerical modeling is an important tool in better understanding past tsunamis and their geophysical sources, in real-time warning and evacuation, and in assessing hazards and mitigating the risk of future tsunamis. I will discuss a variety of techniques from adaptive mesh refinement to probabilistic hazard analysis that are being used for tsunamis and related geophysical hazards.