For more details CSWeek.lbl.gov

For more details CSWeek.lbl.gov

Berkeley Lab – CS Seminar
The Scientific Python project and the Berkeley Open Source Program Office

Date:
02/05/25

Time:
1:30 pm – 2:30 pm

Location:
59-4102 and zoom

Speakers(s) and Affiliations(s):
Jarrod Millman, Scientific Python Developer at BIDS and the Executive Director for Berkeley’s Open Source Program Office (OSPO)

Title:
The Scientific Python project and the Berkeley Open Source Program Office

Abstract:
With an extensive and high-quality ecosystem of libraries, scientific
Python has emerged as the leading platform for data analysis. This
ecosystem has been created and sustained by independent volunteers
with separate mailing lists, websites, roadmaps, documentation,
engineering and packaging solutions, and governance structures.
Unfortunately, this has meant that there was a lack of coordination
that resulted in duplicated effort, disorganized documentation,
breakage upon new releases, unintended performance regressions, and
user confusion.

The Scientific Python project (https://scientific-python.org/)
addresses these issues through a number of efforts:

– Providing recommendation infrastructure for community-wide policy,
– Improving common engineering infrastructure,
– Hosting shared development and community documentation,
– Maintaining a set of classroom-style lecture notes,
– Organizing a series of developer events to discuss needs,
– Writing a community vetted strategic plan, and
– Helping the community develop grant proposals.

Bio:
Jarrod Millman is a Senior Open Source Scientific Python Developer at
BIDS and the Executive Director for Berkeley’s Open Source Program
Office (OSPO). With a background in computer science, mathematics, and
statistics, and degrees from Cornell and Berkeley, Millman is a
founding member of the scientific Python ecosystem. His primary focus
is on developing and sustaining open-source, community-owned
scientific software tools. Millman serves on the steering council of
NetworkX, is a core developer of scikit-image, and was an early
contributor to NumPy, SciPy, and scikit-learn. He has co-founded
several influential initiatives to advance open and reproducible
research, including the Scientific Python project, the nonprofit
NumFOCUS, and the Neuroimaging in Python project.

Virtual Information:
https://lbnl.zoom.us/j/94777801965

Meeting ID: 947 7780 1965

Host of Seminar:
Lavanya Ramakrishnan, Scientific Data Division

For more details CSWeek.lbl.gov

Agenda:
1- minute for safety
1-minute for health and wellness
DC News and Events
Local updates
TBD

Join from a PC, Mac, iPad, iPhone or Android device:
    Please click this URL to join. https://lbnl.zoom.us/j/94610984579?pwd=T1E1dk15ZXdkOWpMVDhPczlzUFlVZz09
    Passcode: 157876

Seminar Type: Computing Sciences Area Seminar

~~~~~~~~Calendar event name~~~~~~~~~~~~~
CS Seminar: Geometric Deep Learning and Generative Modeling for Drug Discovery and 3D Biomolecular Design
~~~~~~~~Event Details Below~~~~~~~~~~~~~~
Berkeley Lab – CS Seminar
Geometric Deep Learning and Generative Modeling for Drug Discovery and 3D Biomolecular Design

Date:
02/06/25

Time:
11:00AM – 12:00PM

Location:
https://lbnl.zoom.us/j/95239048013

Speakers(s) and Affiliations(s):
ALEX MOREHEAD, UNIVERSITY OF MISSOURI-COLUMBIA

Title:
Geometric Deep Learning and Generative Modeling for Drug Discovery and 3D Biomolecular Design

Abstract:
The COVID-19 pandemic underscored the critical need for rapid drug discovery, catalyzing advances in machine learning (ML) for understanding and designing biological molecules. This domain, often referred to as AI for Science, represents a convergence of modern ML with high-impact scientific challenges. This talk will explore the emerging paradigm of integrating geometric deep learning algorithms with generative ML techniques to model dynamic 3D biomolecules. By enabling scalable, rational molecular design for applications such as drug discovery and bioengineering, these approaches simultaneously advance research in ML and biotechnology. In this seminar, I will highlight recent successes, such as advancements in modeling protein-ligand interactions, discuss current interdisciplinary challenges, and explore future directions at the intersection of AI and molecular science.

Bio:

Virtual Information:
Join Zoom Meeting
https://lbnl.zoom.us/j/95239048013

Meeting ID: 952 3904 8013

Berkeley Lab – CS Seminar

Date:
Thursday, February 6, 2025

Time:
3:00pm – 4:00pm

Location:
Shyh Wang Hall – Bldg. 59, Room 3070

Speakers(s) and Affiliations(s):
Yaniv Brick
Ben-Gurion University of the Negev, Beer-Sheva, Israel

Title:
Generalized Integral Equation Formulations for Efficient High-Frequency Wave Analysis

Abstract:
High-fidelity wave simulation of large and intricate systems has been enabled greatly thanks to advances to the numerical solution of integral equation (IE) formulations of scattering problems. Fast solution algorithms and more stable formulations have been progressively developed to combat the high computational costs and degradation in problem conditioning associated with the increase in the electrical size, model detail, and numbers of simulated scenarios. Recent research in these fronts has focused on the development of fast direct solvers, relying on algebraic compression and compressed factorization of impedance matrices. Among these, notable are the classes of first generation low-rank compression-based hierarchical matrix solvers and the more sophisticated butterfly-compression solvers. The latter exhibit superior asymptotic cost scaling, even for geometric settings that do not exhibit inherent (asymptotic) rank-deficiency, i.e., ones that are not of reduced dimensionality. The appeal of the former is in their relative simplicity and effectiveness for systems that include large quasi-planar or elongated parts. Complementary to the search for faster and more compressive algebraic structures is that for IE formulations that are of enhanced compressibility, which will be the focus of our talk (selected publications listed below). Generalized IE formulations use modified Green’s function kernels to turn, where possible, broadside wave interactions between large sub-structures, which are of fast-scaling ranks, into reduced dimensionality ones. The resulting moment matrix blocks exhibit slower scaling of the ranks and greater rank-deficiency. This translates to greater compression and faster compressed factorization and solution. In the talk, we will review the evolution of generalized IEs, discuss the various types of such formulations, and present algorithms for removal of the bottlenecks associated with the computations of their more complex IE kernels. We will then demonstrate the algorithmic utilization of the reduced effective dimensionality (e.g., by employing non-uniform and randomized field sampling technique) for the design of fast solvers.

[1] Y. Brick, V. Lomakin, and A. Boag, “Fast Green's function evaluation for sources and observers near smooth convex bodies,” IEEE Trans. Antennas Propag., vol. 62, no. 6, pp. 3374 – 3378, June 2014.

[2] A. Sharshevsky, Y. Brick, and A. Boag, “Direct solution of scattering problems using generalized source integral equations,” IEEE Trans. Antennas Propag., vol. 68, no. 7, pp. 5512-5523, July 2020.

[3] D. Zvulun, Y. Brick, and A. Boag “A generalized source integral equation for enhanced compression in three dimensions,” IEEE Trans. Antennas Propag., vol. 72, no. 12, pp. 9316-9325, Dec. 2023.

[4] Y. Dahan and Y. Brick , “Fast direct solvers with arbitrary admissibility using generalized source integral equations,” IEEE Trans. Antennas Propag., vol. 72, no. 10, pp. 7872-7882, Oct. 2024.

Bio:
Yaniv Brick received the B.Sc. (2005, magna cum laude), M.Sc. (2007, summa cum laude), and Ph.D (2015) degrees in electrical and electronics engineering from the School of Electrical Engineering, Tel Aviv University. From 2014 to 2017, he was a Post-Doctoral Fellow with the Oden Institute for Computational Engineering and Sciences, The University of Texas at Austin. Since 2017, he has been with the School of Electrical and Computer Engineering, Ben-Gurion University of the Negev, where he is an associate professor. He is currently a visiting scholar at the University of California San Diego Center for Memory and Recording Research (CMRR). He has been an associate editor for the IEEE Journal on Multiscale and Multiphysics Computational Techniques and the IEEE Antennas and Propagation Magazine. He is a senior member of the IEEE and the International Union of Radio Science (URSI), and a recipient of the IEEE AP-S Doctoral Research Award, the Peter O’Donnell Jr. Post-Doctoral

2 Fellowships in Computational Engineering and Sciences, the Fulbright Post-Doctoral Fellowship Program Grant, and the Fulbright Alumni Prize.

Virtual Information:
Join Zoom Meeting
https://lbnl.zoom.us/j/97229812706

Meeting ID: 972 2981 2706

—

One tap mobile
+16692192599,,97229812706# US (San Jose)
+16699006833,,97229812706# US (San Jose)

—

Dial by your location
• +1 669 219 2599 US (San Jose)
• +1 669 900 6833 US (San Jose)
• +1 213 338 8477 US (Los Angeles)
• +1 720 928 9299 US (Denver)
• +1 971 247 1195 US (Portland)
• +1 206 337 9723 US (Seattle)
• +1 253 215 8782 US (Tacoma)
• +1 346 248 7799 US (Houston)
• +1 602 753 0140 US (Phoenix)
• +1 301 715 8592 US (Washington DC)
• +1 312 626 6799 US (Chicago)
• +1 470 250 9358 US (Atlanta)
• +1 470 381 2552 US (Atlanta)
• +1 646 518 9805 US (New York)
• +1 646 558 8656 US (New York)
• +1 651 372 8299 US (Minnesota)
• +1 786 635 1003 US (Miami)
• +1 267 831 0333 US (Philadelphia)

Meeting ID: 972 2981 2706

Find your local number: https://lbnl.zoom.us/u/awbD2wS8A

—

Join by SIP
• [email protected]

—

Join by H.323
• 144.195.19.161 (US West)
• 206.247.11.121 (US East)
• 115.114.131.7 (India Mumbai)
• 115.114.115.7 (India Hyderabad)
• 159.124.15.191 (Amsterdam Netherlands)
• 159.124.47.249 (Germany)
• 159.124.104.213 (Australia Sydney)
• 159.124.74.212 (Australia Melbourne)
• 64.211.144.160 (Brazil)
• 159.124.168.213 (Canada Toronto)
• 159.124.196.25 (Canada Vancouver)
• 170.114.194.163 (Japan Tokyo)
• 147.124.100.25 (Japan Osaka)

Meeting ID: 972 2981 2706

Host of Seminar:
Yang Liu
Scalable Solvers Group (SSG)
Applied Mathematics and Computational Research Division
Lawrence Berkeley National Laboratory