Nachman Group @ Berkeley Lab - Join Us!

Berkeley Undergraduates

We have many projects that are suitable for undergraduate students to make a significant research contribution. There is a very strong track record of our students publishing papers as part of their research in the group:

• Vishnu Sangli ('24), arXiv:2305.17169
• Norman Karr ('22), ML4PS@NeurIPS 2022
• Sulaiman Alvi ('23), JHEP 02 (2023) 220
• Kasia Krzyzanska^* ('22), JHEP 01 (2023) 061
• Jerry Lai ('22), JHEP 04 (2022) 156
• Jason Huang ('21), JINST 16 (2021) P05001
• Rebecca Hicks ('21), Phys. Rev. A 103 (2021) 022407 and Phys. Rev. A 105 (2022) 012419
• Adi Suresh ('22), Phys. Rev. D 103 (2021) 036001 and ICLR simDL (2021) 12
• Kees Benkendorfer and Luc Le Pottier^* ('22), Phys. Rev. D 104 (2021) 035003
• Josh Lin ('21), JHEP 05 (2019) 181 and JHEP 10 (2018) 101
• Harley Patton ('19), Nucl. Instrum. Meth. A 913 (2018) 91
• Alex Spies ('18), JINST 14 (2019) P09028
• Sam Bright-Thonney ('18), JHEP 03 (2019) 098
• Yiten Chen ('17), Nucl. Instrum. Meth. A 902 (2018) 197

^*Visiting Berkeley students from Princeton (Kasia), Reed (Kees), and U. Michigan (Luc).

During the school year, the typical path to joining our group is through the URAP program. You can apply to our project Deep Learning Methods for High Energy Physics. If you have any programming and/or machine learning experience, please mention this. If you have never taken a statistics/statistical learning/machine learning class or have no programming experience, please consider taking classes in these topics before applying. Berkeley has many excellent classes in both of these areas. If you are unsure which classes to take, please feel free to send Dr. Nachman a brief message and he would be happy to offer advice.

In exceptional cases, Dr. Nachman will also consider students who write him an email directly. Please do not write if you have not taken a statistics/statistical learning/machine learning class or have no programming experience. If you want to increase the chances that he will respond in the affirmitive, please try running an example notebook from any of the recent papers we have written and mention this in the email. You can find links for the code in the publications tab.

During the semester, we can arrange for course credit through URAP. During the summer, we have funds for paid internships. The best way join our group during the summer is to join the group in the spring before through one of the mechanisms discussed above.

Berkeley Graduate Students

If you are interested in a 3-6 month project to learn about the interface of machine learning and fundamental physics, please send Dr. Nachman an email. A number of Berkeley (e.g. on rotation) and visiting students have made significant contributions to projects in this way! Dr. Nachman can also serve as a thesis adviser to UC Berkeley graduate students - please get in contact if you are intersted in learning more about this possibility. Contact information for current students can be found here.

If you are looking to build a strong machine learning foundation, there are many great resources available to you! The first place to look is the set of courses at your institution. At Berkeley, the intro graduate sequence is statistics is STAT 200A/B. This should provide a strong formal foundation. If you already have some background, I understand that stats-oriented ML graduate students take 205a/205b (might be too formal), 210a/210b (the first one is probably more relevant), and 241a/241b. I would also not be afraid of "undergraduate" courses like STAT 154 and others. There is also the exciting course Physics 288 (ML for physics), but it does not replace a solid set of courses from the statistics or computer science departments.

There are also many online courses which can provide a quick introduction and often are more applied. Andrew Ng's course is perhaps the most popular. I understand that the course from Haslie and Tibshirani (also on edX) is excellent, as well as the one from Abu-Mostafa. I have not carefully looked through any of these courses. If you take them, please let me know if they are useful!

For reference, here is a list of graduate students that have worked closely with Dr. Nachman for a signficant portion of their Ph.D. research:

• Matan Grinberg (UC Berkeley 2nd year), topic TBD
• Radha Mastandrea (UC Berkeley 2nd year), Anomaly Detection with Machine Learning
• Krish Desai (UC Berkeley 3rd year), Machine Learning-based Cross Section Measurements
• Murtaza Safdari (Stanford University '22), Thesis.
• Aviv Cukierman (Stanford University '20), Thesis, in particular the search for new particles using weak supervision (published in PRL).
• Veronica Wallängen (Stockholm University '19), Thesis, in particular the modeling of silicon radiation damage (published in JINST).
• Rebecca Carney (Stockholm University '19), Thesis, in particular the modeling of silicon radiation damage (published in JINST).
• Jennifer Rolloff (Harvard University '19), Thesis, in particular the measurement of jet mass (published in PRL).
• Michela Paganini (Yale University '19), Thesis, in particular the Calorimeter Generative Adversarial Network (published in PRL).
• Zihao Jiang (Stanford University '19), Thesis, in particular the measurement of gluon splitting to bottom quarks (published in PRD).
• Christian Herwig (University of Pennsylvania '19), Thesis, in particular the measurement of interference in events with b-quarks and leptons (published in PRL).

Prospective LBNL Postdocs

• Machine Learning-based Unfolding: aprecruit

Here are older searches that have now been closed:

• Inference-aware Machine Learning: Taleo
• Machine Learning for Particle Identification: Taleo
• Machine Learning Anomaly Detection: AJO
• 2023 Chamberlain Fellowship: AJO
• Developing Surrogate Models for Particle Physics and Cosmology: AJO
• Machine Learning Detector Design and Data Analysis for Hadronic Physics: AJO
• NESAP for Learning Fellow: INSPIRE
• 2021 Chamberlain Fellowship (deadline Oct. 16, 2020): AJO

Please send me an email with questions!

Prospective Visitors

There are many opportunites to visit LBNL and work with our group, both during the semester as well as during the summer. Here are some possible funding sources:

• Science Undergraduate Laboratory Internships (SULI). This is great for undergraduate summer research (similiar to REU) or postbac research during the school year.
• Office of Science Graduate Fellowship (SCGF) Program. This program provides funding for graduate students to be based at Berkeley Lab.
• US ATLAS Center. This program provides funding for students and postdocs to work with researchers at the national laboratories. Proposals are relatively lightweight but have to be coordinated with lab staff.

Please send me an email with questions or proposals (including if you apply to SULI).