NA Digest, V. 21, # 45

NA Digest Sunday, December 05, 2021 Volume 21 : Issue 45


Today's Editor:

Daniel M. Dunlavy
Sandia National Labs
dmdunla@sandia.gov

Today's Topics: Subscribe, unsubscribe, change address, or for na-digest archives: http://www.netlib.org/na-digest-html/faq.html

Submissions for NA Digest:

http://icl.utk.edu/na-digest/



From: Luis Nunes Vicente lnv@lehigh.edu
Date: December 02, 2021
Subject: DFO-TR, Derivative-free Optimization Solver


10 years ago, Katya and I wrote a Matlab code for continuous
optimization of functions without using their derivatives. The code
was called DFO-TR and has been used by colleagues and collaborators.

DFO-TR runs a trust-region interpolation based method. It is
essentially described in Section 5 of the paper A. S. Bandeira,
K. Scheinberg, and L. N. Vicente, Computation of sparse low degree
interpolating polynomials and their application to derivative-free
optimization, Mathematical Programming, 134 (2012) 223-257.

We are now releasing DFO-TR to the community (please see
https://coral.ise.lehigh.edu/lnv/dfo-tr). Version 0.1 (November 2021)
is written in Matlab and solves small, unconstrained DFO problems
efficiently and robustly.

DFO-TR is freely available for research, educational or commercial
use, under a GNU lesser general public license.

DFO-TR team:
Liyuan 'Leon' Cao (Peking University)
Tommaso Giovannelli (Lehigh University)
Katya Scheinberg (Cornell University)
Oumaima Sohab (Lehigh University)
Luis Nunes Vicente (Lehigh University)



From: Kris ONeill oneill@siam.org
Date: December 03, 2021
Subject: New Book, Modern Nonconvex Nondifferentiable Optimization


Modern Nonconvex Nondifferentiable Optimization
by Ying Cui and Jong-Shi Pang

Starting with the fundamentals of classical smooth optimization and
building on established convex programming techniques, this research
monograph provides a foundation and methodology for modern nonconvex
nondifferentiable optimization by providing readers with theory,
methods, and applications of nonconvex and nondifferentiable
optimization in statistical estimation, operations research, machine
learning, and decision making. A comprehensive and rigorous treatment
of this emergent mathematical topic is urgently needed in today's
complex world of big data and machine learning. This book takes a
thorough approach to the subject and includes examples and exercises
to enrich the main themes, making it suitable for classroom
instruction.

SIAM Bookstore:
https://my.siam.org/Store/Product/viewproduct/?ProductId=3D39767459

November 2021 / xx + 756 pages / Hardcover / 978-1-611976-73-1 /
List $119.00 / SIAM Member $83.30 / MO29



From: GEORGE ANASTASSIOU ganastss2@gmail.com
Date: December 03, 2021
Subject: New Book, Unification of Fractional Calculi with Applications


UNIFICATION OF FRACTIONAL CALCULI WITH APPLICATIONS
GEORGE A. ANASTASSIOU

In this monograph we demonstrate the unifying methods of generalized
versions of Hilfer, Prabhakar and Hilfer-Prabhakar fractional calculi
and we establish related unifying fractional integral inequalities of
the following types: Iyengar, Landau, Polya, Ostrowski,
Hilbert-Pachpatte, Hardy, Opial, Csiszar's f-Divergence, self-adjoint
operator and related to fuzziness. Our results are univariate and
multivariate. This monograph's results are expected to find
applications in many areas of pure and applied mathematics, especially
in fractional inequalities and fractional differential
equations. Other interesting applications can be in applied sciences
like geophysics, physics, chemistry, economics and engineering. This
monograph is appropriate for researchers, graduate students,
practitioners and seminars of the above disciplines, also to be in all
science and engineering libraries.

https://www.springerprofessional.de/en/unification-of-fractional-calculi-with-applications/19887148




From: Michael Dumbser michael.dumbser@unitn.it
Date: December 03, 2021
Subject: Numerical Methods for Hyperbolic PDE Course, ONLINE, Jan 2022


Short Course on Advanced Numerical Methods for Hyperbolic Equations
Laboratory of Applied Mathematics, University of Trento, Italy
The course will also be made available online via ZOOM

Lecturers: Prof. Dr.-Ing. Michael Dumbser and Dr. Firas Dhaouadi
Dates: January 31st - February 4th 2022
Website: http://www.unitn.it/nm2022
Course fee: 250 EUR (online), other fees for onsite participation

This advanced short course is primarily designed for PhD students and
post-doctoral researchers in applied mathematics, scientific
computing, computational physics and computational mechanics.

Summary: The course consists in a structured intensive one-week
programme of 40 hours of theoretical lectures and computer laboratory
exercises on advanced numerical methods for hyperbolic partial
differential equations with applications in engineering and science.
The course covers high order finite volume and discontinuous Galerkin
methods, also on unstructured triangular simplex meshes; Riemann
solvers, higher order TVD, ENO and WENO schemes, the discretization of
hyperbolic PDE with non-conservative products, well balanced schemes
as well as meshless Lagrangian particle methods. The lectures are
supplemented with many computer laboratory exercises to provide
practical hands-on experience concerning the practical aspects of the
implementation of these advanced numerical methods.

Further information can be found on the dedicated web page
http://www.unitn.it/nm2022

The deadline for registration is January 30th 2022.

For further questions, please contact:
Prof. Dr.-Ing. Michael Dumbser, michael.dumbser@unitn.it



From: Pedro Valero-Lara valerolarap@ornl.gov
Date: December 03, 2021
Subject: Extreme Heterogeneity Solutions, South Korea, April 2022


ExHET 2022: The 1st International Workshop on Extreme Heterogeneity
Solutions to be held in conjunction with PPoPP 2022, April 2-5, 2022,
Seoul, South Korea

CFP links:
https://excl.ornl.gov/ppopp-exhet-2022/
https://easychair.org/cfp/ExHET22

Submission link:
https://easychair.org/my/conference?conf=3Dexhet22#

While computing technologies have remained relatively stable for
nearly two decades, new architectural features, such as specialized
hardware, heterogeneous cores, deep memory hierarchies, and
near-memory processing, have emerged as possible solutions to address
the concerns of energy-efficiency, manufacturability, and
cost. However, we expect this 'golden age' of architectural change to
lead to extreme heterogeneity and it will have a major impact on
software systems and applications. In this upcoming exascale and
extreme heterogeneity era, it will be critical to explore new software
approaches that will enable us to effectively exploit this diverse
hardware to advance science, the next- generation systems with
heterogeneous elements will need to accommodate complex
workflows. This is mainly due to the many forms of heterogeneous
accelerators (no longer just GPU accelerators) in this heterogeneous
era, and the need of mapping different parts of an application onto
elements most appropriate for that application component.



From: Pamela Bye pam.bye@ima.org.uk
Date: December 03, 2021
Subject: Mathematics of Finance, UK, Jun 2022


8-10 June 2022, Holiday Inn, Liverpool, UK
https://ima.org.uk/17964/ima-conference-on-mathematics-of-finance-and-climate-risk/

This conference on financial mathematics will be focusing on pressing
challenges in finance and insurance produced by climate change,
demographic developments, and the ever-increasing dominance of data
and information. Alongside the challenges, this three-day conference's
goal will devote time to the role and responsibilities financial and
insurance mathematics have in developing solutions and in assisting
with transitions necessary to mitigate irreversible adverse
environmental and socio- economic impact. The aim is to promote
interdisciplinary cooperation bridging mathematics, statistics and
computer science with finance, climate science, insurance and
economics. Industry professionals and members of government agencies
will be invited to share their experience and expertise alongside
academic experts in the field in order to scope out the most urgent
research directions with the highest potential for incisive solutions.

Talks will be accepted for the conference based on a 350 word
abstract. Abstracts should be submitted by 15th January 2022 either
online at http://online.ima.org.uk or by e-mail to
conferences@ima.org.uk. Authors will be notified by 15th February
2022 if their abstract has been accepted for oral or poster
presentation. Oral presentations are expected to be 25 minutes in
length, including time for questions and answers. Abstracts should
include: Whether your title is intended for oral or poster
presentation; Title of the talk; Authors and Affiliations; Intended
Speaker; 350 words describing talk.

Registration is currently open:
https://my.ima.org.uk/services.php?section=3Devents

For scientific queries please contact: Paul Johnson,
Paul.Johnson-2@manchester.ac.uk . For general conference queries
please contact Maya Everson, Conference Officer E-mail:
conferences@ima.org.uk .




From: Ronnie Sircar sircar@princeton.edu
Date: December 01, 2021
Subject: Assistant Professor Position, ORFE, Princeton Univ


The Department of Operations Research and Financial Engineering (ORFE)
at Princeton University invites applications for a tenure-track
faculty appointment at the Assistant Professor level starting
September 1, 2022. The search is in the area of Optimization &
Operations Research, connecting with ongoing and planned strategic
initiatives in the School of Engineering and Applied Science. In
particular: HealthTech (for instance, optimal vaccine rollout
strategies); Energy and the Environment (for instance, integrating
renewable energy production, and optimizing the electricity grid of
the future); Robotics and Cyberphysical Systems (uncertainty
quantification, safety verification, and joint learning and control of
dynamical systems); Resilient and Smart Cities (how smart cities could
use information technologies for efficient deployment and utilization
of perishable resources); and Data Science (for instance, optimization
of machine learning algorithms). A PhD in a related field is required.

To be successful, the candidate must have a strong commitment to
excellence in research and in teaching at both the undergraduate and
graduate levels. The ORFE department believes that the diversity of
our faculty, staff, and students is essential to the distinction and
excellence of our research and academic programs. To that end, we are
eager to have a colleague who supports our institutional commitment to
ensuring Princeton is inclusive, equitable, and diverse.

The ORFE department is part of the School of Engineering and Applied
Science and involved in activities with the Center for Statistics and
Machine Learning, the Bendheim Center for Finance, the Program in
Applied and Computational Mathematics, and the Andlinger Center for
Energy and the Environment. An appointment may be made jointly with
another department or program.

Applications will be considered on a continuing basis, but candidates
are encouraged to apply by December 15, 2021. To apply, please submit
an online application at
https://www.princeton.edu/acad-positions/position/23641. All
applicants should include a CV, research statement, teaching
statement, and contact information for at least three references, one
of whom should be able to address the candidate's teaching abilities.



From: Sarah Dance s.l.dance@reading.ac.uk
Date: December 03, 2021
Subject: Postdoc Position, Data Assimilation, Univ of Reading


The Department of Meteorology at the University of Reading seeks
applications for a post-doctoral research role which will provide an
exciting opportunity for the post-holder to contribute to and develop
research work on observation impact in hazardous weather prediction:
measuring the ability of different observation types to improve
forecasts of hazardous weather through data assimilation. The post is
jointly funded by the University of Reading and the UK Met Office.

In a changing climate, an improved ability to forecast hazardous
weather is key to the management of risk for society. Observations
play an essential role in numerical weather prediction but are
expensive to obtain. Quantitative measures of the impact of
observations on weather forecasts allow evaluation of the best use of
currently available observations and the design of future observing
networks, to help ensure that observational data is used in the most
cost-effective way for the benefit of society and the economy.
Recently, international collaborative research efforts have developed
tools for quantifying the observation impact on global weather
forecasts. However, these tools are not applicable to the next
generation of regional hazardous numerical weather prediction systems
due to nonlinearity and statistical sampling issues. The Met Office
Next Generation (NG) system is expected to include high resolution
ensemble forecasts, driven by high frequency, spatially dense
observation datasets. New approaches are needed to deal with these
large data volumes that take advantage of modern data science. The
post-holder will co-create their research with scientists at the
University of Reading and the UK Met Office.

For further information and to apply see
https://jobs.reading.ac.uk/displayjob.aspx?jobid=3D8884
Deadline for applications 4 January 2022.



From: HRUSHIKESH N MHASKAR hrushikesh.mhaskar@cgu.edu
Date: December 01, 2021
Subject: Postdoc Position, ML/Signal Processing, Claremont Graduate Univ


A postdoc position is available to work on an NSF grant project with
Hrushikesh Mhaskar (CGU, https://www.cgu.edu/people/hrushikesh-
mhaskar/). The areas of research are (1) approximation
theory/computational harmonic analysis (2) machine learning, and (3)
signal processing. The post-doc will collaborate closely with Mhaskar
in (1) developing applications for his theory and writing the
necessary codes to carry out the applications successfully enough to
result in publications in respectable journals/proceedings (2) doing
literature search and grant proposal development, (3) mathematical
discussions and presentations. The ideal candidate should have good
programming skills, and a background in basic real and complex
analysis, approximation theory, harmonic analysis, and numerical
analysis. The deadline for applications on mathjobs
(https://www.mathjobs.org/jobs?joblist-2718-17968) with cover letter,
vita, statement of research interests, and three letters of
recommendation is December 15, 2021, but applications will be
considered after that until the position is filled. The position is
from July 1, 2022, to June 30, 2023, but the dates are somewhat
flexible. Claremont Graduate University is a member of the Claremont
Colleges, a consortium of two graduate and five undergraduate
institutions that collectively house over 50 mathematics faculty
members. The individual institutions cooperate, through the Claremont
Center of Mathematical Sciences, to form one of the largest
mathematical science research communities in California. The Institute
of Mathematical Sciences runs a PhD program in Mathematics as well as
joint PhD programs with other universities in Southern California. In
addition, there are often clinics sponsored by local
industries. Postdocs are eligible for on campus housing:
https://www.claremontcollegiateapartments.com/



From: Marta D'Elia mdelia@sandia.gov
Date: December 01, 2021
Subject: Postdoc Position, Sandia Labs/CA, USA


A postdoctoral position is available at Sandia National Laboratories
in Livermore, California, starting immediately. We are looking for
creative problem solvers with knowledge of machine learning, deep
learning, and, more generally, with experience in scientific
computing, evidenced by publications and codes. Knowledge and
experience in Bayesian statistics and uncertainty quantification are
also desirable. The work mainly involves the design of new machine
learning algorithms for the prediction of material structure and
properties, for given physical process conditions. It also includes
the design and implementation of efficient optimal experimental design
methods, in both deterministic and Bayesian settings. The position
will involve extensive interaction with a large and diverse project
team, code development and analysis, and demonstration of the
developed technology in multiple applications. This project is focused
on discovering new resilient materials and manufacturing processes via
an artificial-intelligence-guided approach.

Qualifications We Require: PhD in applied mathematics, computational
science or engineering; Knowledge and expertise in machine learning
and neural networks; Knowledge and expertise in computational science
and code development. Qualifications We Desire: Knowledge of
optimization, Bayesian statistics, and uncertainty quantification;
Expertise in Python and C++; Experience in performing collaborative
research; Excellent communication skills.

To apply for this job, please, go to the Sandia's career page:
https://www.sandia.gov/careers/career-possibilities/students-and-postdocs/internships-co-ops/postdoctoral-positions/
and click on "Livermore", then, select the position with Job Opening
ID: 679662, and Posting Title: Postdoctoral Appointee - Machine
Learning in Material Science. For more information, please, contact
Marta D'Elia at mdelia@sandia.gov



From: David Hyde david.hyde.1@vanderbilt.edu
Date: December 03, 2021
Subject: Postdoc Positions, Physical Simulation + Learning, Vanderbilt Univ


One or more postdoctoral positions are available in David Hyde's group
in the Department of Computer Science at Vanderbilt University in
Nashville, TN, USA. The postdoctoral scholar will work on projects at
the intersection of computational physics, computer graphics,
learning, vision, applied mathematics, high-performance computing,
etc. (there is flexibility depending on the candidate's background and
interests). Potential research topics include simulating new or
existing physical (e.g. solid/fluid) phenomena at exascale; developing
new formulations for solid-fluid coupling; building
vision/learning-based models for incorporating real-world data into
physical simulations and thereby improving simulation accuracy; using
learning techniques to power higher-fidelity real-time
simulations/effects; etc. Please see the group website for examples
of past projects, and please see the job posting for a full
description and application. Candidates are reviewed on a
first-come/rolling basis.

Group website: https://dabh.io

Job posting:
https://www.vanderbilt.edu/postdoc/position-detail/?id=3D602



From: Daniel Appel=F6 appeloda@msu.edu
Date: December 01, 2021
Subject: PhD Positions, Computational Math/Sci/Eng, Michigan State Univ


Michigan State University's Department of Computational Mathematics,
Science and Engineering (CMSE; https://cmse.msu.edu/) is currently
accepting applications for its interdisciplinary PhD program in
computational and data science. This program provides its graduates
broad and deep knowledge of the fundamental techniques used in
computational modeling and data science, significant exposure to at
least one application domain, and the opportunity to conduct
significant original research in algorithms and/or applications
relating to computational and data science.

A brief introduction to the program can be found at here, and a list
of CMSE faculty and their research interests can be found at
http://cmse.msu.edu/faculty. Additional information about the PhD
program, as well as application information, can be found at
http://cmse.msu.edu/apply. Please note that the deadline for
application for the Fall 2022 cohort is January 2, 2022.

Please forward this message along to any students that you think might
be interested in our graduate program.


End of Digest
**************************