April 12, 2024
The BAND team working on ROSE (the Reduced Order Scattering Emulator) has just published a paper describing ROSE in Physical Review C. The paper explains in detail the theory and computational framework behind the ROSE software, as well as demonstrating it’s ability to enable Bayesian Uncertainty Quantification of the parameters that appear in nuclear optical potentials.
March 22, 2024
On March 18 BAND member Özge Sürer gave an invited talk on the BAND Framework as part of the STAR collaboration’s “Spring Juniors Day”.
February 19, 2024
BAND Principal Investigator, Daniel Phillips, is spending Spring 2024 at the Chalmers University of Technology in Gothenburg, Sweden. He is a Tage Erlander Visiting Professor and so is on leave from his position at Ohio University. The work he will conduct during the Visiting Professorship is strongly tied to the goals of BAND, as you can read in this article.
January 20, 2024
Three BAND PhD students received their doctorate in Fall Semester. Dan Liyanage has joined PayPal as a Machine Learning Scientist, after earning his Physics doctorate at Ohio State. Mookyong Son started a position as a Statistics Research Fellow at Unlearn.AI, having completed his PhD in Statistics and Probability at Michigan State. And Moses Chan, who’s now working as an Assistant Professor of Instruction in the Engineering School at Northwestern, finished his PhD in Industrial Engineering and Management Sciences at Northwestern. Congratulations to Dr Chan, Dr Liyanage, and Dr Son!
January 10, 2024
Two BAND post-docs have recently moved into permanent positions! In December Pablo Giuliani transitioned to being a Specialist at FRIB/MSU, focused on Nuclear Science and Graduate Student Success, while also conducting research in Bayesian analysis and Machine Learning for nuclear physics. A few months earlier Daniel Odell joined Savannah River National Laboratory as a Data Scientist.
December 11, 2023
BAND members Pablo Giuliani and Alexandra Semposki, as well as BAND Fellow Jason Bub, presented their research at the recent joint meeting of the American Physical Society Division of Nuclear Physics and the Physical Society of Japan. All three talks were part of the Mini-Symposium on “Advanced Statistics and Machine Learning Methods in Nuclear Physics”. Giuliani helped to organize the Mini-Symposium and gave an invited talk during it.
November 12, 2023
BAND member Witek Nazarewicz and two statistician collaborators, Vojta Kejzlar and Léo Neufcourt, co-authored a paper, published this week in Scientific Reports. In it, they propose a Bayesian statistical machine learning framework that uses the Dirichlet distribution to combine results of several imperfect models. They show that global and local mixtures of models reach excellent performance on both prediction accuracy and uncertainty quantification and are preferable to classical Bayesian model averaging. See the BAND highlight for more details.
October 18, 2023
BAND members Moses Chan, Özge Sürer, Stefan Wild, and John Yannotty attended the INFORMS annual meeting in Phoenix, AZ, October 15 - 18. During the meeting, Wild received the 2023 Egon Balas Prize from the INFORMS Optimization Society and shared his insight in the designing of algorithms for derivative-free optimization problems. Chan, Sürer, and Yannotty each presented recent development in computational methods to facilitate uncertatinty quantification, in a two-part session “Computational Methods for Uncertainty Quantification”, organized by Chan. Details can be found under the list of BAND methodology presentations.
October 10, 2023
BAND has released v0.3 of our software framework! This release includes a number of new tools, including: BMEX, a web application for exploring nuclear masses and related quantities; parMOO, a parallel multiobjective simulation optimization library; rose, a reduced-order scattering emulator; and Taweret, a package containing multiple Bayesian Model Mixing methods. It also includes updates to the surmise and SAMBA packages. Pull the repo to try these new capabilities and tell us about your experience!
September 28, 2023
Congratulations to BAND co-PI Stefan Wild (LBNL and Northwestern), for being awarded the 2023 Egon Balas Prize! The Egon Balas Prize of the INFORMS Optimization Society was established in 2020 and is awarded annually for a body of contributions in the area of optimization. The award ceremony will take place at the INFORMS annual meeting on October 15 in Phoenix, AZ.
August 14, 2023
The third BAND annual retreat has successfully concluded! The retreat was held from August 8 - 9 at the Ohio University Dublin Integrated Education Center. During the retreat, BAND members gathered to review the progress made for various projects and share exciting updates in software development. Prior to the retreat on August 7, the external Advisory Board met with BAND members. We greatly appreciate the Board for their advice and support for BAND!
June 29, 2023
BAND members Pablo Giuliani, Kyle Godbey, Frederi Viens, and Alexandra Semposki co-organized the FRIB-TA Summer School on Practical Uncertainty Quantification and Emulator Development in Nuclear Physics. The school ran from June 26th to June 28th at the Facility for Rare Isotope Beams on the campus of Michigan State University.
June 9, 2023
Postdoc Sunil Jaiswal has joined the BAND team! Sunil received his Ph.D. from the Tata Institute of Fundamental Research in Mumbai, India. He will be working at the Ohio State University with Dick Furnstahl, Ulrich Heinz, and Matt Pratola.
May 26, 2023
ISNET-9, held at Washington University in St. Louis from May 23 to 26, 2023, featured presentations by BAND researchers: Ozge Sürer, Dan Liyanage, Witek Nazarewicz, Dick Furnstahl, Pablo Giuliani, Daniel Odell, and Kyle Godbey, and former BAND Fellow Jason Bub. Slides are available at the ISNET-9 website.
May 22, 2023
BAND Camp 2023 was a great success! It was held just before ISNET-9 at Washington University in St. Louis on May 22, 2023, and featured interactive presentations (with Python code via Google Colab) by BAND researchers Matt Pratola, Moses Chan, Ozge Sürer, Daniel Odell, Pablo Giuliani, Alexandra Semposki, John Yanotty, Kyle Godbey, Witek Nazarewicz, and Dick Furnstahl. Slides and videos are available here.
May 16, 2023
Double congratulations to BAND graduate student Moses Chan! First off, the paper he wrote “Constructing a simulation surrogate with missing output” together with fellow BAND members Matt Plumlee and Stefan Wild, was published earlier this month in Technometrics. Second, in Fall 2023 Moses will join the faculty at Northwestern as an Assistant Professor of Instruction; he will spearhead their Data Science & Engineering program.
April 17, 2023
The Bayesian R-matrix Inference Code Kit (BRICK), which is part of the BAND software framework, has recently been used by BAND member Daniel Odell, his collaborator James deBoer, and their experimental colleagues to analyze several nuclear reactions of astrophysical interest. Particularly notable is their Nature paper on the 19F(p, γ)20Ne rate in primordial stars. Odell and deBoer collaborated with scientists from the China JinPing Underground Laboratory to analyze new data on this reaction taken at that facility. They found its rate in the first stars was roughly seven times higher than previously thought. This would mean enhanced breakout from the ‘warm’ carbon-nitrogen-oxygen cycle, and could explain previously puzzling observations of calcium in the first stars. Further experimental data from underground laboratories, observations from the James Webb Space Telescope, and analyses using Bayesian methods will help refine this picture further.
March 6, 2023
BAND member Dick Furnstahl and co-authors Christian Drischler, Jordan Melendez, Alberto Garcia, and Xilin Zhang have two recent publications that present an overview of model order reduction methods for the creation of fast & accurate emulators and a pedagogical introduction to projection-based, reduced-order emulators for applications in low-energy nuclear physics. Stay tuned for the upcoming release of the Reduced Order Scattering Emulator software we are developing at BAND!
February 8, 2023
This year’s BAND camp will take place at Washington University, St. Louis, on Monday, May 22nd. The BAND collaboration is sponsoring accommodation for the BAND camp for up to forty participants. And, as a special offer to viewers of this channel, if you attend the BAND camp we’ll also pay for you to stay on for the ISNET (Information and Statistics in Nuclear Experiment and Theory) Workshop that immediately follows it! Visit the ISNET-9 meeting website for more details and to register. But register soon! Remember: the early registrant gets the free room!
January 9, 2023
Edgard Bonilla, Kyle Godbey, and BAND post-doc Pablo Giuliani are co-authors on two recent publications that employ Reduced Basis Methods (RBM) to speed up the calibration and evaluation of density functionals. In Training and projecting: a reduced basis method for many-body physics they and collaborator Dean Lee lay out the foundation of the method and demonstrate its potential in some illustrative examples. In Bayes goes fast: Uncertainty quantification for a covariant energy density functional emulated by the reduced basis method Bonilla, Godbey, and Giuliani teamed up with BAND member Frederi Viens and Jorge Piekarewicz to calibrate a relativistic energy density functional using a fraction of the computational resources that would have been needed without RBM.
December 5, 2022
Congratulations to BAND member Stefan Wild! Stefan took up a new position as division director of Lawrence Berkeley National Laboratory’s Applied Mathematics and Computational Research Division on December 1.
November 19, 2022
BAND member Witek Nazarewicz and his 17 co-authors have just published a Reviews of Modern Physics Colloquium on Machine Learning in Nuclear Physics. You can also read a nice summary of their paper in this phys.org piece.
November 5, 2022
Five BAND members–co-PI, Özge Sürer, post-docs Pablo Giuliani and Daniel Odell, graduate student Dan Liyanage, and BAND leader Daniel Phillips–gave presentations at the recently concluded 2022 Fall Meeting of the Division of Nuclear Physics of the American Physical Society. You can check out what they all said at our Presentations page.
October 24, 2022
The paper Interpolating between small- and large-g expansions using Bayesian model mixing written by BAND members Alexandra Semposki, Dick Furnstahl, and Daniel Phillips has been published in Phys. Rev. C. It tests BAND’s linear, bivariate, and multivariate model mixing approaches on a toy model and paves the way for applications to crucial nuclear physics problems in the very near future!
October 14, 2022
BAND has released v0.2 of our software framework! This release includes two “BAND tools” intended to facilitate Bayesian analyses: surmise for model emulation & calibration and SaMBA for model mixing. It also includes three “BAND examples” where we apply these tools and methods to forefront Nuclear Physics problems: BFRESCOX for coupled-channel analyses of nuclear reactions, BRICK for R-matrix calculations, and a tutorial (QGP_Bayes) on the use of JETSCAPE_SIMS tools to analyze the quark-glion plasma. Pull the repo to get all of them, or pull some of it to get particular pieces you want, but please go ahead and play and tell us what you think when you do!
Setpember 13, 2022
A guide to the use of Order-Reduction Methods in Nuclear Physics, written by Jordan Melendez, BAND co-PI Dick Furnstahl, Christian Drischler, Alberto Garcia, and Xilin Zhang, has been published in J. Phys. G.
July 17, 2022
The annual BAND Retreat will be held July 21 and 22 in Dublin, Ohio. Details (schedule, directions, etc.) are available here.
May 1, 2022
Congratulations to BAND post-doc, Özge Sürer! She will be joining the faculty at Miami University. No, not that one, the one in Oxford. No, not that Oxford, the one in Ohio. Dr Sürer will be taking up a position as an Assistant Professor of Business Analytics in August.
February 28, 2022
Multiple members of BAND participated in the NSF Project Scoping Workshop Towards Precise and Accurate Calculations of Neutrinoless Double Beta Decay. The workshop took place on January 31 and February 1 over Zoom, and involved statisticians, computer scientists, and physicists working on calculations of neutrinoless double beta decay of nuclei. It was organized by Jon Engel, from the University of North Carolina, together with BAND members Witek Nazarewicz and Daniel Phillips. A white paper summarizing the conclusions of the workshop is in preparation.
January 26, 2022
We are very proud to anounce that two BAND team members, Taps Maiti and Filomena Nunes, were named 2021 Fellows of the American Association for the Advancement of Science. Congratulations on their distinguished contributions in statistics and nuclear physics -both foundation pillars of BAND-, as well as their continuous excellence and engagement in teaching the next generation of scientists!
December 17, 2021
The Second Annual BAND camp was a lot of fun. We had 47 particpants, 21 of whom attended in person. Talk .pdfs can be downloaded here. Ozge Surer’s excellent tutorial on BFRESCOX is available on github.
December 7, 2021
Join the BAND team! The Ohio State University has a post-doc opening. The successful candidate will work effectively in an inter-disciplinary team to build a sustainable code base that can solve problems at the statistics-physics interface. The position will focus on Bayesian modeling and methodology development for mixing multiple physics simulation models of nuclear dynamics, and building scalable computational algorithms for inference, prediction and uncertainty quantification.
November 30, 2021
The schedule for the Second Annual BAND Camp is now set! We are looking forward to pedagogical presentations on BAND research by Frederi Viens, Ozge Surer, and Jordan Melendez. BAND leader Daniel Phillips will preview BAND’s “comming attractions”. And we’ll hear talks by recent Ph.D.s in this area who now work in the insurance industry.
November 26, 2021
The paper Precision mass measurement of light self-conjugate nucleus 80Zr by BAND team member Witek Nazarewicz and his experimental and theoretical collaborators has been published in Nature Physics. The journal selected the paper as a highlight pointing in particular to the use of one of the techniques at the heart of BAND, Bayesian Model Averaging.
September 13, 2021
Post-doc Daniel Odell will be joining the BAND team at Ohio University on October 1! Daniel has been a postdoc at OU since summer 2019 after completing his Ph.D. at the University of Tennessee under the direction of Lucas Platter.
August 18, 2021
Check out BAND-supported research on emulators: Fast & accurate emulation of two-body scattering observables without wave functions and Toward emulating nuclear reactions using eigenvector continuation.
August 2, 2021
The ISNET 8 conference will be held in hybrid mode on December 14-16, 2021, preceded on December 13 by pedagogical lectures at the Second Annual BAND Camp. All talks will be delivered on-site, but sessions will also be broadcasted via zoom for remote participants. For more information and to register, see the ISNET 8 page.
June 22, 2021
The 2021 BAND Retreat was held on June 21 jointly on zoom and with in-person contingents in Michigan and Ohio. It was great to see everyone (some in person
for the first time in over a year) and to hear all the progress!
June 16, 2021
The external BAND Advisory Board met with BAND members on June 14 via zoom. Our thanks to the Board for their advice and support!
May 20, 2021
The BAND Manifesto, Get on the BAND Wagon: A Bayesian Framework for Quantifying Model Uncertainties in Nuclear Dynamics, is now published in Journal of Physics G. Full text here.
February 6, 2021
We’ve released surmise. surmise is a Python package that’s designed to provide a surrogate model interface for calibration, uncertainty quantification, and sensitivity analysis. An example is provided with the package so please go ahead and get in and play! Comments & questions to Ozge Surer who, together with Moses Chan, Matt Plumlee and Stefan Wild, developed this very useful tool.
January 1, 2021
Post-doc Ozge Surer joined the BAND team! Ozge completed her Ph.D. at Northwestern University under the supervision of Dan Apley and Ed Malthouse in December 2020. She will be working at the Northwestern-Argonne Institute of Science & Engineering under the supervision of Matt Plumlee and Stefan Wild.
January 1, 2021
Post-doc Pablo Giuliani joined the BAND team! Pablo completed his Ph.D. at Florida State University under the direction of Jorge Piekarewicz in December 2020. He will be working at Michigan State University with a joint appointment in the Department of Statistics and Probability under the supervision of Frederi Viens and the Facility for Rare Isotope Beams.
December 18, 2020
The First Annual BAND Camp was a great success! Daniel Phillips (OU) gave an overview of BAND, Michael Grosskopf (LANL) reviewed “Bayesian Basics”, Simon Mak (Duke) and Derek Everitt (OSU) explained “Model Emulation and Parameter Estimation”, and Matt Pratola (OSU) went through “Bayesian Analysis of Nuclear Dynamics: Early Directions in Bayesian Model Mixing”. The slides for the BAND Camp as well as for the Virtual ISNET 8 workshop that followed can be found here.
December 14, 2020
The BAND Manifesto, entitled Get on the BAND Wagon: A Bayesian Framework for Quantifying Model Uncertainties in Nuclear Dynamics, is online on arXiv. Check it out!
September 24, 2020
Registration for Virtual ISNET 8 and for the First Annual BAND Camp is now open. Information and registration can be found here. The website also lists confirmed speakers for both events.
September 17, 2020
Opening for Post-doctoral Research Associate in BAND collaboration at Ohio University (see online ad to apply). Review begins November 24, 2020.
August 26, 2020
Paper on Efficient emulators for scattering using eigenvector continuation is online in Physics Letters B 809, 135719 (2020) [also on arXiv].
July 1, 2020
The BAND Framework project officially begins! Press releases from Ohio University, The Ohio State University, and Michigan State University.