Department of Mathematics

# Colloquium Series

## Upcoming **Colloquia**

Colloquia will resume in Fall 2024. Have a great summer!

**Recent Colloquia**

**Friday, April 19, 2024 - Symeon Papadimitropoulos, Ph.D., and Nathan Willis, Ph.D.
(UC Merced postdocs)**

**At 9:00 AM in PB 103**

**Title: **Synthetic aperture imaging using physically informed convolutional neural networks
(Dr. Papadimitropoulos)

**Abstract: **We propose a physically informed neural network (PINN) approach to address the problem
of locating scatterers using synthetic aperture radar (SAR). The proposed approach
uses the physical properties of wave propagation to aid the neural network in learning
to detect and localize scatterers from SAR data. This methodology is especially efficient
in super-resolution cases, i.e. when the scatterers are located close to each other,
at distances smaller than the nominal resolution of the imaging system. Using the
proposed approach, we are able to detect an a priorly unknown number of scatterers
and localize them with high accuracy.

**Short Speaker Bio: **Symeon is a Postdoctoral Scholar at the University of California, Merced. Prior to
that, he held postdoctoral appointments at Tel-Aviv University and The Technion. Symeon
obtained his Ph.D. degree in 2018 from the University of Crete, in Greece. His research
interests include absorbing boundary conditions, numerical methods for PDEs, inverse
problems relating to wave propagation and imaging, and underwater acoustics. Current
projects involve the incorporation of deep learning methods in these areas.

**Title: **Shallow-water simulations of colliding turbidity currents (Dr. Willis)

**Abstract:** Motivated by future deep-sea mining for polymetallic nodules, we investigate colliding
turbidity currents in one and two horizontal dimensions. The standard shallow-water
equations are amended to include the sediment concentration and pressure difference
with the ambient fluid. Our simulations show that currents do not mix upon collision
and instead develop a reflecting wave in each current. We describe the effects of
the ratio of current concentrations, ratio of current heights, settling speed, and
collision angle. To understand the overall impact of the mining process on the ocean
floor, we present the resulting patterns of sediment deposition.

**Short Speaker Bio:** Nathan Willis is a Visiting Assistant Professor at the University of California,
Merced. In May 2022, Nathan graduated with his PhD in Mathematics under the advisement
of Christel Hohenegger. Nathan utilizes numerical approximations, asymptotic analysis,
and applied analysis to research a variety of applied problems arising from fluid
dynamics. Some examples of ongoing research includes surface tension in free surface
sloshing, steady streaming in non-Newtonian fluids, and colliding turbidity currents
arising in deep sea mining. As a member of Project NExT, Nathan is also passionate
about teaching mathematics and incorporating innovative and active learning techniques
in his classroom.

**Friday, April 5, 2024 - **

**At 9:00 AM on Zoom (Note: Pre-registration is required.)**

**Title:** Using Supercomputers to Understand Our World

**Friday, March 1, 2024 - Bethany Lusch (Argonne National Laboratory)**

**At 9:00 AM on Zoom (Note: Pre-registration is required.)**

**Title:** Accelerating Simulations with Machine Learning

**Abstract:** Large-scale simulations are used in many fields to address questions such as how
to design an efficient engine, how climate change will affect a particular region,
or how cancer spreads in the body. There are always more simulations that would be
useful to run than available time, especially when each simulation takes weeks on
a supercomputer. Can we leverage data saved from previous simulations to speed up
future ones? In this talk, I will cover a couple of ways that we have used machine
learning to accelerate simulations.

**Short Speaker Bio:** Dr. Bethany Lusch is an Assistant Computer Scientist at Argonne National Lab. She
is on the data science team within Argonne's supercomputing facility. Her work focuses
on machine learning for large-scale scientific problems. Her postdoctoral training
was in the area of machine learning for dynamical systems at the University of Washington
in Seattle. She obtained her PhD and MS in applied mathematics at the University of
Washington. Dr. Lusch also holds a BS degree in mathematics from the University of
Notre Dame.

*Celebrating Women in Math Series*

**Tuesday, March 5, 2024 - Abigayle Dirdak, PhD Candidate (University of Arizona)**

**At 4:00 PM in PB 032**

**Title: **How would you describe a Mathematician? Oh, and some stuff about the Pentagram map!

**Abstract: **Have you ever pondered about what defines a mathematician? Join us while we discuss
the topic. Additionally I will be presenting on my research involving polygons, projective
spaces, and something called the pentagram map. So if polygons and/or discussions
on mathematicians interest you, come to my talk.

**Short Bio:** Abigayle grew up on a small family farm in central California. After graduating high
school she attended College of the Sequoias before transferring to Fresno State. Participating
in research projects with Dr. Carmen Caprau, and Dr. Khang Tran inspired her to continue
doing math and to attend graduate school. Abigayle received her BA in Mathematics
from Fresno State in 2018, and is now excited to be finishing her final year as a
PhD student in mathematics at the University of Arizona.

**Friday, March 8, 2024 - Jennifer Elder, PhD (Missouri Western State University)**

**At 10:00 AM in S2 207**

**Title: **Finding Needles in Haystacks

**Abstract: **Picture a one way street with n parking spots. Parking preferences are tuples of numbers
that correspond to n cars' preferred parking spots on the street. For example, (1,3,1)
would mean there are three spots, car one wants to park in spot 1, car two wants to
park in spot 3, and car three also wants to park in spot 1. A preference tuple is
called a parking function if all the cars are able to park in their preferred spot,
or some spot further down the street. In this talk, we will discuss a special class
of parking functions called unit interval parking functions. These are the parking
functions in which cars park at most one spot away from their preferred parking spot.
We will discuss a variety of new bijections, recursions, and generating functions
produced by working with unit interval parking functions, including solutions to previously
unsolved problems from Richard Stanley.

**Short Bio:** Jennifer earned both a Bachelor’s and a Master’s Degree in Mathematics at Fresno
State. Her Master’s Thesis was on “Generalizations of the Futurama Theorem,” a problem
related to cycle decompositions of permutations. After graduating in 2016, she went
on to earn a PhD in Mathematics from Arizona State University in 2021. Her Dissertation
research was in Combinatorics, also on permutation properties. After her PhD, she
spent two years as a Visiting Professor of Mathematics at Rockhurst University in
Kansas City Missouri, and was a Postdoc TA for the 2022 Summer@ICERM Undergraduate
Research Program at Brown University. As of Fall 2023, she is an Assistant Professor
of Mathematics at Missouri Western State University in Saint Joseph Missouri. She
is focused on teaching General Education Math classes, and is an active researcher
in Combinatorics.

**Thursday, November 16, 2023 - Ayse Ozturk, PhD (The Ohio State University, Newark)**

**At 1:00 PM in PB 390 and Zoom**

**Title: **Bridging Access and Equity with Mathematical Modeling and Teacher Preparation

**Abstract:** Dr. Ozturk will outline her overarching research agenda, encompassing access and
equity in mathematics education, mathematical modeling, and teacher preparation. She
will highlight the key findings from her last two publications and offer insights
into the future direction of her ongoing research initiatives.

**Short Bio: **Ayse Ozturk is a lecturer of Mathematics Education in the Department of Teaching and
Learning and the Department of Mathematics at The Ohio State University, Newark. Her
research focuses on the learning and teaching of mathematics as a humanizing practice.
As a teacher educator, she is dedicated to crafting equity-based lessons that empower
students to become active and critical mathematical thinkers.

**Tuesday, November 16, 2023 - Casey Griffin, PhD Candidate (University of Delaware)**

**At 1:00 PM in PB 390 and Zoom**

**Title: **Women’s Sense of Belonging in Undergraduate Calculus

**Abstract: **Prior research has identified low sense of belonging as a key reason why women leave
undergraduate STEM majors, especially after taking Calculus. Historically, Calculus
instruction has been primarily lecture-based, though recently, efforts have shifted
toward incorporating instruction that provides opportunities for students to engage
in active learning. Prior studies have suggested that incorporating active learning
may support students’ sense of belonging, however, there lacks consensus on the particular
types of active learning opportunities that best support students. In my talk, I will
be presenting findings from two studies investigating women’s sense of belonging in
undergraduate Calculus, and the influence of the learning opportunities they experience
during class on their sense of belonging. My results suggest that women tend to report
positive influences of active learning opportunities on their sense of belonging,
particularly those that allow them to interact with their classmates and with their
instructor. I conclude by describing directions where I hope to take this work.

**Short Bio:** Casey Griffin is a Ph.D. candidate specializing in mathematics education at the University
of Delaware’s School of Education. Casey is interested in learning more about the
ways in which active learning opportunities in undergraduate mathematics courses might
support underrepresented students’ sense of belonging, and further serve as a way
to diversify the STEM major population. Her dissertation focuses on women’s sense
of belonging in an undergraduate Calculus course that incorporates frequent opportunities
for students to engage in active learning, and how students perceive the impact of
these opportunities on their sense of belonging. Before attending graduate school,
Casey earned a B.A. in Secondary Mathematics Education from the University of Delaware
and taught high school mathematics in Long Branch, New Jersey for two years. She then
went on to pursue a M.A. in Mathematics from Villanova University and returned to
the University of Delaware for her Ph.D.

**Wednesday, November 8, 2023 - Ernesto Daniel Calleros, PhD Candidate (San Diego State
University, UC San Diego)**

**At 2:00 PM in PB 390 and Zoom**

**Title:** Investigating the Language Demands and Resources of InquiryBased Mathematics Education
for Linguistically Diverse Students: The Case of Inquiry-Oriented Linear Algebra

**Abstract:** Prior research shows that inquiry-based instructional approaches are more effective
than lectures. However, these approaches might not yield equal benefits for students
from certain marginalized groups (e.g., see Reinholz et al., 2022). Given the centrality
of communication in inquiry-based mathematics classrooms, one important group to consider
is linguistically diverse students whose primary language may differ from the language
of instruction. In this presentation, I explore an overarching question related to
equity for linguistically diverse students: What language demands and resources do
linguistically diverse students experience in an inquiry-oriented linear algebra course?
Grounded in a situated sociocultural theory of learning, I constructed a framework
that captures language demands and resources along three interrelated dimensions:
(a) lexico-grammatical – words, phrases, and grammar, (b) situational – material,
activity, semiotic, and sociocultural aspects, and (c) normative – social and socio-mathematical
norms. The data analyzed were classroom observations of one inquiry-oriented linear
algebra course taught by an expert instructor, as well as semi-structured interviews
with 5 male linguistically diverse students from the course. The 5 students spanned
a diverse range of cultural backgrounds (Korean, Vietnamese, Malaysian, Hispanic,
and White) and comfort levels with English. The data was analyzed using thematic analysis
(Miles et al., 2020), allowing for both a priori and emergent codes. Results indicated
that language demands experienced by linguistically diverse students included: unclear
meanings of mathematical notation and terminology, verbal dominance, and violated
expectations about a single “right” way to solve a task. Language resources included:
student familiarity with problem contexts, group roles based on different language
systems, and translation tools. These findings shed light on the language challenges
and resources that linguistically diverse students may experience in inquiry-based
mathematics classes. I conclude with recommendations for addressing the identified
linguistic challenges and leveraging the identified linguistic resources.

**Short Bio:** Ernesto Daniel Calleros is a Ph.D. candidate in the Mathematics and Science Education
joint doctoral program at San Diego State University and the University of California
San Diego. Mr. Calleros earned an M.A. in Mathematics from Rice University and a B.S.
in Applied Mathematical Sciences from Texas A&M University. His research interests
include undergraduate mathematics education and investigating ways to make mathematics
linguistically accessible to all students, especially multilingual learners, in K-16
educational contexts. Mr. Calleros has taught mathematics in both college-level and
K-12 settings.

**Friday, November 3, 2023 - Chris Barker, Ph.D. (Statistical Consulting Section of
the American Statistical Association) )**

**At 9:00 AM in PB 390 and Zoom **

**Title: **Sample Size for Number of Patient Interviews when Developing a PRO by the 2009 FDA
Guidance

**Abstract: **Patient Reported Outcomes (PRO’s) developed according to the 2009 FDA PRO guidance
require an initial step of structured patient interviews or focus groups for “item
Generation”. The 2009 guidance does not provide sample size suggestions or methodology
for number of interviews (sample size) for item generation. This talk defines a “Potential”
Type I error, an URN model for use in simulation, simple descriptive statistics and
pragmatic methodology of Kaplan Meier and “capture recapture” for estimating a sample
size for the Item Generation step in the 2009 FDA PRO guidance. These methods also
applicable in the more recent update to the 2002 “Patient-Focused Drug Development:
Selecting, Developing, or Modifying Fit-for-Purpose Clinical Outcome Assessments “
(PFFD ) guidance. This work proposes an URN model for use in simulating the Item Generation
process, and the adoption of a Kaplan Meier statistical methodology and additional
Capture- Recapture sample size methodology that address this gap in the FDA PRO guidance.
This work also appears to be the first to provide a definition of potential Type I
error in interviews for a PRO methods for assessing sample size and a new descriptive
statistics to summarize saturation based on a probability distribution to confirm
whether enough interviews have been prepared. A Potential Type I error is declaring
saturation when it has not been achieved during the interviews. Two worked examples
applied to actual interview data and published data, are presented. Guidelines are
proposed for the estimation of sample size useful for PRO experts conducting interviews
for a PRO. These methods are applied in the setting of qualitative research. Future
research for other methods of interviewing and determining saturation is required
as well as incorporating the correlation among items during the item generation process.

**Short Bio: **Chris Barker is a Consultant Biostatistician, an Adjunct Associate Professor of Biostatistics
at the University of Illinois, and in the process of retiring. Chris’ background,
training and Ph.D. are in Bio-statistics from the University of Illinois Graduate
School of Public Health. He has worked in the pharmaceutical industry on drug and
vaccine development, medical devices and disease diagnostics. Chris has also worked
in a Global Health Economics and Strategic Pricing group for a large international
pharmaceutical company (Roche) collaborating with economists developing cost-benefit,
cost-utility, and cost consequences for new drug development. During the pandemic,
Dr. Barker has been working on new drugs for covid19 and medical diagnostics for detection
of SARS-COV-2, the virus causing covid19. His work in the pharmaceutical industry
has permitted him to give lectures around the world including London UK, Basel Switzerland,
Suva Fiji and Havana Cuba. One of his current interests is in use of “role play” for
training students and statisticians throughout their career in Statistical Consulting
Skills.

**Friday, October 13, 2023 - Elliott Vest, PhD Candidate (UC Riverside) (Fresno State
alumnus)**

**At 9:00 AM in PB 390**

**Title:** The relevance of CAT(0) cube complexes in Geometric Group Theory

**Abstract:** Geometric Group Theory, as the name suggests, is the study of groups from a geometric
perspective. The hope is to associate a group with a metric space, then use the geometry
of the metric space to find more information about the group. A very popular collection
of metric spaces we use is CAT(0) cube complexes, formed by gluing collections of
Euclidean cubes together via isometries of their faces. The relatively easy geometry
of CAT(0) cube complexes has contributed to the resolutions of long-standing conjectures,
and they have shown to be a fruitful tool in the study of mapping class groups. This
talk starts with an overview of Geometric Group Theory, then pivots to how CAT(0)
cube complexes are defined as well as what makes their geometry so nice. With this
foundation, we will see how recently created tools in the field of Geometric Group
Theory can extend arguments for CAT(0) cube complexes to a larger class of spaces;
namely CAT(0) spaces, hierarchically hyperbolic spaces, or even proper geodesic metric
spaces.

**Short Bio:** As of Fall 2023, Elliott Vest is a PhD mathematics student finishing his last year
at University of California, Riverside. Born and raised in Visalia, California, he
began his college journey at the College of the Sequoias in Visalia, then transferred
to Fresno State in 2017 as a mathematics major. He graduated from Fresno State in
2019. It was the inspiration of many of the Fresno State faculty, such as Dr. Comlan
De Souza and Dr. Katherine Kelm, that led him to apply to a PhD program in mathematics
in the first place. From tutoring, to facilitating labs at Fresno State, to now teaching
mathematics at UC Riverside as well as mentoring incoming graduate students, Elliott
has a strong passion for teaching and hopes to obtain a teaching position after graduation.
He enjoys going to his family cabin at Bass Lake, playing piano, going to the gym,
destroying opponents at ping pong, and roleplaying in Dungeons and Dragons with his
friends.

**Friday, October 6, 2023 - Youngsoo Choi (Lawrence Livermore National Laboratory)**

**At 9:00 AM by Zoom - Preregister (using QR code in the flyer) to receive the link.**

**Title: **Physics-constrained data-driven methods for accurately accelerating simulations

**Abstract: **See the flyer

**Friday, May 5, 2023 - Joseph Sifuentes, Ph.D. (UTRGV) **

**At 3:30 PM in PB 011**

**Title: **Why I Became a Mathematician - A Tale of Art, Heavy Metal, Eigenvalues and Killer
Muscle Cars

**Abstract: **In 2011, President Obama awarded Applied Mathematics Professor Richard Tapia the National
Medal of Science for his seminal contributions to the field of optimization and the
tremendous work he has done as a mentor and advocate for young mathematicians from
underrepresented groups. In this talk, I'll tell you how he did that using a customized
1970 Super Sport Chevelle named Heavy Metal and the prospect of making a music video.
This was how I began my irreversible descent into a love affair with mathematics and
research.

**Short Bio: **Dr. Josef Sifuentes is a boot-loving Texan who majored in Math, Computational and
Applied Mathematics, and Art as an undergraduate at Rice University and stayed to
complete a Ph. D. in Computational and Applied Mathematics there. He has worked as
a research scientist at the Courant Institute at NYU and a visiting professor at Texas
A&M before joining the faculty in the School of Mathematical and Statistical Sciences
at UTRGV. His research interests are in numerical linear algebra and wave scattering
problems. Dr. Sifuentes has served as a mentor to talented students as director of
the UTRGV Pathways to Math Graduate School, co-director of the NSF-LSAMP summer research
academy and faculty mentor for the Gulf States Math Alliance, amongst other programs
aimed at promoting undergraduate research involvement. And he really likes BBQ and
coffee.

**Friday, April 28, 2023 - Alex Kontorovich, Ph.D. (Rutgers University)**

**At 11 AM by Zoom only**

**Title:** (Potential) Applications of Machine Learning to Pure Mathematics

**Abstract:** Recent developments both in Machine Learning and Interactive Theorem Prover technology
lead to interesting and important questions about the future of research mathematics.
We will discuss some of these advances and speculate on potential changes and challenges
in the medium and perhaps even short term horizon.

**Short Bio:** Kontorovich is a Professor of Mathematics at Rutgers University. He received his
BA in Mathematics from Princeton, followed by a PhD in Mathematics from Columbia.
Before moving to Rutgers, Kontorovich taught at Brown, Stony Brook, and Yale. Kontorovich’s
research interests span number theory, dynamical systems, ergodic theory, geometry,
and representation theory. He is a Fellow of the American Mathematical Society, a
2017 Kavli Fellow of the National Academy of Sciences, and during the 2020-21 year,
he was also Distinguished Visiting Professor for the Public Dissemination of Mathematics
at the National Museum of Mathematics (MoMath) in New York.

**Friday, April 21, 2023 - Dr. Nicolle Gonzalez (UC Berkeley)**

**At 12 PM in PB 013**

**Title: **Crystals!

**Abstract:** A basic question in linear algebra is finding the coefficients of a vector in terms
of a given basis. A fundamental goal in combinatorics is finding formulas for these
coefficients when our vector spaces are certain polynomial rings. It turns out questions
like these can be encoded and answered using certain directed graphs known as crystals.
However, these crystals contain much more information than the original polynomials
themselves. Algebraic properties like symmetries of the polynomials and multiplication
rules are all encoded as symmetries and tensor product structures on these graphs.
Hence, the polynomials can be seen as "shadows" of the crystals. In fact, the crystals
themselves are also shadows of more complicated algebraic objects known as representations.
In this talk we will introduce crystals from scratch and with lots of examples. We
will show how the product of symmetric polynomials follows from tensoring certain
symmetric crystal graphs. We will also discuss how certain bases in the full polynomial
ring correspond to truncations of these symmetric crystals and how multiplying these
truncated graphs turns out to be unexpectedly mysterious.

**Short Bio: **Nicolle was born and raised in Venezuela. After some time in Hawaii and Oregon for
undergrad, she moved to LA to pursue grad school at USC. She has had postdocs at UCLA
and MSRI and is currently a Morrey visiting professor at UC Berkeley.

**Thursday, March 30, 2023 - Dr. Melissa Zhang (postdoc at UC Davis)**

**At 12 PM in PB 390**

**Title: **Using Algebra to Study the 4D Behavior of Knots

**Abstract: **A knot is a closed loop of string in 3-space. When subject to gentle wiggling and
deforming, a knot does not lose its most inherent properties, making it a foundational
object in many areas of topology. The simultaneous visual intuitiveness and rich diversity
of knots allows them to also serve as bookkeeping tools for many other fields, including
statistical mechanics, algebraic combinatorics, quantum computing, and many more.

Given two random pictures of complicated knots, it would be near impossible for a human to immediately determine whether the knots are different. For tasks like this, we use knot invariants, which assign diagrams of knots to some other mathematical object that we understand better. For instance, a famous knot invariant called the Jones polynomial assigns to each knot diagram a (Laurent) polynomial, and if given two different diagrams of the same knot, it will produce the same polynomial.

Knot invariants like the Jones polynomial are useful not only for telling knots apart but also for classifying them based on their similarities. A more recent enhancement of the Jones polynomial called Khovanov homology assigns a more complex algebraic object to a knot diagram. By studying knots through their Khovanov homology, one can now capture relationships between knots, or more precisely, the evolution of knots throughout 4D spacetime.

In this talk, we will get a taste of how Khovanov homology extends the Jones polynomial to the fourth dimension, and explore several ways the 4D behavior of knots makes them even more interesting.

**Short Bio: **Melissa Zhang is currently a Krener Assistant Professor at UC Davis. Previously, she
was a postdoc at MSRI/SLMath, and the University of Georgia. She earned her PhD at
Boston College under the supervision of her advisors Eli Grigsby and David Treumann.
She is a low-dimensional topologist who studies visualizations of physical objects
like knots, 3D spaces with portals, as well as 4D analogues of knots. In practice,
if you ever watch her doing research, she is probably doing one of the following activities:
(1) drawing pictures, (2) collaborating with friends, (3) trying to write down her
proofs carefully, or (4) thinking through arguments using linear algebra.

She's always loved science and math growing up, but her pure math career really began when she took her first proof-based course in college and loved it! She also enjoys singing, knitting, hiking, and playing with her pet rats.

**Friday, March 24, 2023 - Dr. Scott Mitchell (Sandia Lab)**

**At 9 AM via Zoom (Register in advance at https://us06web.zoom.us/j/83723613079)**

**Title:** Computing geometry as a mathematician in an engineering laboratory

**Abstract: **Geometry is all around us. Sandia creates physical objects with physical shape, and
we need to explore the abstract high-dimensional spaces that describe the variety
of environmental conditions these objects might encounter. They

must always "work" when we want them to, and never "work" when we don't! I study computational
geometry, which is the math and algorithms for calculating geometric quantities of
objects, or creating new objects with prescribed geometry. My favorite work is joining
up with people with a different background from me, such as mechanical engineering,
optimization, uncertainty quantification, geophysics, or data sciences. As a mathematician,
they often ask me *how* to solve some problem they are stuck on. Instead, I ask them
*why* they want to solve it and usually end up redefining their problem so it is easy
to solve. My favorite problems involve the interplay between continuous shapes and
the discrete connections binding them together. I study triangular and cubical meshes,
and how to optimize them for increased simulation accuracy. I also study point-sampling
of abstract spaces, and computational topology to characterize the structure of fragments.
I'll describe my journey from my family to the labs, what it's like to work there,
and work with the international research community. I'll mention a few research highlights
from my nearly 30 years at Sandia and 75 publications. The talk is intended to be
accessible to STEM undergraduates.

**Short Bio: **Scott A. Mitchell received a B.S in Applied Math, Engineering & Physics from the University
of Wisconsin-Madison in 1988. He received an M.S. (1991) and Ph.D. (1993) in Applied
Math from Cornell University. Since Oct 1992 he has been at Sandia National Laboratories.
He researched triangular and tetrahedral meshing algorithms via a computational geometry
approach from 1992-1993. He was part of the Cubit project, doing mesh generation R&
D from 1993-2000, and project leadership from 2000-2002. He managed the Optimization
and Uncertainty Estimation department from 2002-2007. He served on Sandia's LDRD (internal
research program) team and NNSA's ASC program. He moved on to technical work in 2007.
He published about 75 papers, and served on the program committee for the International
Meshing Roundtable and Symposium on Computational Geometry He is a member of SIAM
and ACM.

**Friday, February 24, 2023 - Dr. Kelly Moran (Los Alamos National Lab)**

**At 9 AM via Zoom**

**Title: **Understanding the universe on large scales

**Abstract: **In this seminar, I introduce myself and my background and talk about a cosmology project
I've been working on for a little over a year. You'll hear about how when I first
started working at Los Alamos National Lab (LANL) I knew nothing about statistics,
compared to now when I have a PhD in it! In a similar vein, you'll hear that when
I first started working on this project I knew nothing about cosmology, but now I
know enough to be useful. Cosmology aims to understand the universe on large scales.
In order to maximally extract information from modern cosmological surveys, matching
theoretical predictions are 39 needed. At low redshifts cosmological simulations are
the main way of getting predictions, but their computational cost makes it impossible
to run very large ensembles. We built an emulator for the matter power spectrum based
on the Mira-Titan Universe simulation suite that is designed to be useful for

cosmological inference studies on a range of cosmological probes. It is more accurate
(at the 2-3% level) than competing methods, and covers a larger range of cosmological
parameters.

**Short** **Bio:** Dr. Moran works on developing, analyzing, and visualizing Bayesian statistical models
across varied applications such as global and environmental health, cosmology, and
space physics. She has experience in developing novel methods for probabilistic dimension
reduction and fast Gaussian process approximation. She first joined Los Alamos National
Laboratory as a post-baccalaureate student with the Information Systems and Modeling
(A-1) group, returned throughout graduate school to work with the

Statistical Sciences (CCS-6) group, and hired on with CCS-6 as a staff scientist in
2021. She enjoys being outside, blues and fusion dancing, playing the oboe, and hanging
out with her two large orange cats.

*Celebrating Women’s History Month*

*Celebrating Women’s History Month*

**Friday, March 3, 2023 - Dr. Catherine Cavagnaro (Sewanee: The University of the South)**

**At 11 AM in SII 206**

**Title: **Mathematical Models for Aircraft Longitudinal Motion

**Abstract: **Longitudinal motion for an aircraft refers to forward and vertical translational motion
as well as rotation about the lateral axis. We will present and explore mathematical
models for such motion and extract techniques pilots can use to more safely fly their
aircraft.

**Short Bio:** In addition to teaching mathematics at Sewanee, Catherine Cavagnaro holds FAA Flight
Instructor and Airline Transport Pilot certificates. She teaches aerobatics at the
airport on campus and is a monthly columnist for AOPA Pilot Magazine. In 2018 Catherine
was inducted into the Tennessee Aviation Hall of Fame and in 2022, she was inducted
into the National Flight Instructor Hall of Fame. She enjoys flying with her two
sons around the United States and delights in finding mathematical applications within
aviation.

**Friday, February 3, 2023 - Dr. Gi-Sang Cheon, Sungkyunkwan University**

**At 10 AM in PB 138**

**Title: **Introduction to Riordan arrays and applications

**Abstract: **Enumeration problems in combinatorics arise in many areas of mathematics, notably
in algebra, probability theory, topology, and geometry, as well as in its many application
areas. Although primarily concerned with finite systems, some combinatorial questions
and techniques can be extended to an infinite (specifically, countable) but discrete
setting. A Riordan array (or Riordan matrix) is an infinite lower triangular matrix
encoding the coefficients of certain sequences of power series. Such an array can
be regarded as a vast generalization of Pascal’s triangle. These concepts lead to
an elegant, unified, and fruitful approach for dealing with a host of enumerative
problems.

**Short Bio:** Gi-Sang Cheon is a Korean mathematician working in combinatorial matrix theory, enumerative
combinatorics, and Riordan group theory. He is currently the director of the Applied
algebra and Optimization Research Center (AORC) , SRC-NRF of Korea since 2016. He
is also a professor of the Department of Mathematics at Sungkyunkwan University. He
received his Ph.D. from Sungkyunkwan University in Feb 1991. His research interests
include linear and multilinear algebra, graph theory, and their applications.

**Thursday, December 1, 2022 - William Godoy, PhD (Oak Ridge National Laboratory (ORNL))**

**To attend the talk you will first need to register in advance at https://us06web.zoom.us/j/88406041576**

**Title:** Pursuing graduate school and career paths for PhDs

**Abstract: **In this presentation we will talk about career paths for those planning to pursue
a graduate school degree. We will provide exposure to the audience from my experiences
on different career paths environments: academia, national labs and industry. In particular,
we would point out the importance of having a diverse network for collaboration, being
part of a scientific community, and the role of professional societies. We will also
discuss the different kind of expectations as people progress in their professional
career levels and uptake on early, mid and senior roles. Overall, we hope students
will get exposure to these topics to get a grasp of the landscape in science and help
them navigate these environments.

**Short Bio:** Dr. William F Godoy joined Oak Ridge National Laboratory in 2016, currently a Senior
Computer Scientist in the Computational Science and Mathematics Division. A native
of Peru, his background is in Mechanical Engineering with a strong research and development
focus on the computational aspects of large-scale scientific modeling and simulation
using high performance

computing (HPC) He obtained his PhD in 2009 from the University at Buffalo, The State
University of New York. Prior experiences include a Senior Software Engineer position
at Intel Corporation, 2012-2016, and a postdoctoral fellow at NASA Langley Research
Center, 2009-2012. https://www.ornl.gov/staff-profile/william-f-godoy

**Thursday, November 3, 2022 - Earvin Balderama, Ph.D. (Fresno State)**

**By Zoom at 12 PM**

**Title:** Statistical Models for Count Data

**Abstract:** Count data are a type of data that takes on non-negative integer values {0, 1, 2,
...}. The Poisson and negative binomial distributions are commonly used as the basis
in a generalized linear model for such data. The negative binomial is used in situations
where the Poisson fit is inadequate due to over-dispersion. When there is an excess
amount of zeros in the data, zero-inflated models and hurdle models are used, again
typically specified by Poisson and negative binomial. What if negative binomial is
still inadequate? In this talk, I will discuss the double-hurdle model, an extension
of the single-hurdle model, to describe count data that contain both an excessive
amount of zeros and some extremely large counts that cause so much over-dispersion
that even a negative binomial specification is inadequate. The double-hurdle model
was applied to counts of marine birds observed along the US Atlantic coast. A Bayesian
MCMC framework was used for estimation and validation. An R package, hurdlr, was created
to implement hurdle and double hurdle models to zero-inflated data.

**Friday, October 28, 2022 - Nathan Urban, Ph.D. (Brookhaven National Laboratory (BNL))**

**By Zoom at 9 AM**

**Title:** PROJECTING THE FUTURE: QUANTIFYING UNCERTAINTIES IN HOW THE CLIMATE WILL CHANGE

**Abstract: **As carbon dioxide emissions to the atmosphere continue from fossil fuel consumption,
the Earth's climate will change considerably over the course of this century due to
the resulting greenhouse effect. Societies and ecosystems will feel these global changes
in the form of extreme weather, recurrent flooding and droughts, and other regional
impacts. But how certain are we in projections about the distant future? Every prediction
of a physical theory comes with error bars arising from limitations in data and approximations
that are made for the sake of computational tractability. The scientific question
is not whether the climate will change, but how likely it is to change by a given
amount. This talk will discuss at a conceptual level the general mathematical formalism
of statistical uncertainty quantification, and how it is being applied within the
U.S. Department of Energy and other research institutions to estimate the plausible
range of possible climate futures. The mathematical methods discussed may include
Bayesian parameter estimation for nonlinear regression, Monte Carlo sampling, sensitivity
analysis, model averaging, and surrogate modeling or emulation of expensive computer
simulations. Applications may include probabilistic predictions of global warming
and climate feedbacks, sea level rise from Antarctic ice sheet disintegration, and
coastal flooding, among others. As time permits, I will also discuss educational training
and career paths into this research field at the forefront of interdisciplinary science.

**Short Bio: **Dr. Nathan Urban leads the Applied Mathematics group in the Computational Science
Initiative at Brookhaven National Laboratory, on Long Island, New York. He received
undergraduate degrees in physics, computer science, and mathematics from Virginia
Tech, and a Ph.D. and M.Ed. in physics (computational statistical mechanics) from
Penn State. After graduating, he moved into climate uncertainty quantification with
postdoctoral appointments at Penn State and Princeton, and a staff position at Los
Alamos National Laboratory. He received a DOE Office of Science Early Career Research
award for multi-model climate uncertainties, and has led major projects involving
coastal resilience planning under uncertainty and "in-situ" methods for embedding
scalable statistical inference algorithms within exascale simulations. At Brookhaven
he develops methodologies for uncertainty quantification, decision making under uncertainty
and optimal experimental design, model reduction, scientific machine learning, and
integrated computational frameworks for decision support, applied to problems in climate
science, biomedicine, materials science, and others.

**Friday, October 7, 2022 - Dr. Mario Bencomo**

**In PB 013 at 3 PM**

**Title:** Topics in inverse problems

**Abstract:** *Inverse problems* are a classification of mathematical problems in engineering and science that involve
estimating model parameters/inputs relative to a “forward/direct” problem. In this
talk I will present several examples of inverse problems to elucidate the definition
and highlight typical mathematical challenges. I will also discuss in depth applications
in the field of exploration seismology where such inverse problems require mathematical
tools from numerical optimization and differential equations.

**Short Bio:** Mario J. Bencomo joined the Department of Mathematics at Fresno State as an Assistant
Professor this Fall semester. Prior to coming to California, he was a Pfeiffer Postdoctoral
instructor at Rice University, in the Department of Computational Applied Mathematics
and Operations Research (CMOR), where he received his Ph.D. in 2017. His research
interests include numerical methods for wave propagation problems, inverse problems,
optimal control with a focus on nonlinear conservation laws.

**If you need a disability-related accommodation or wheelchair access information, please
contact the Mathematics Department at 559.278.2992 or e-mail ** mathsa@csufresno.edu. **Requests should be made at least one week in advance of the event.**