Thematic Einstein Semester on

Mathematical Optimization
for Machine Learning

Summer Semester 2023

New: Proceedings deadline extended to January 15, see below

The semester is organized within the framework of the Berlin Mathematics Research Center MATH+ and supported by the Einstein Foundation Berlin. We are committed to fostering an atmosphere of respect, collegiality, and sensitivity. Please read our MATH+ Collegiality Statement.

Scope of the Semester

The Thematic Einstein Semester Mathematical Optimization for Machine Learning  aims at unlocking the potential of mathematics within the extremely large and diverse fields of study that constitute modern Computer Science and Data Science. It is intended to bring together young researchers and experienced scholars from mathematics and other disciplines and will consist of specific events, continuous activities over the semester, and research visits.

Opening Day (February 10 at Technische Universität Berlin)

There will be an official announcement of the Thematic Einstein Semester along with a reception open to all MATH+ members after the MATH+ Friday talk of Stephen J. Wright on February 10 at the Technische Universität Berlin.

Workshop:  Exploring synergies: Machine Learning meets Physics and Optimization (April 26-28 at Zuse Institute Berlin)

The first workshop of the Thematic Einstein Semester will be held from April 26 to 28 at the Zuse Institute Berlin and focus on bridging the gap between Optimization and Machine Learning. In particular it will cover topics relating to the use of Machine Learning in Physical Systems and Optimization as well as aspects relating to Discrete Optimization.
More details can be found on the workshop homepage.

Workshop on Optimization (June 14-16 at Humboldt-Universität zu Berlin)

The second workshop of the Thematic Einstein Semester will be held from June 14 to 16 at the Humboldt-Universität zu Berlin and focus on optimization methods and applications. In particular it will cover topics relating to the use of Optimal Control, advancements in First-order Methods, Multilevel Optimization and Computer-assisted proofs.

More details to follow can be found on the workshop homepage.

Joint TES/GAMM CoMinds Workshop (July 12-14 at Humboldt-Universität zu Berlin)

The third workshop of the Thematic Einstein Semester will be held from July 12 to 14 at the Humboldt-Universität zu Berlin. It is organized together with the GAMM Activity group Computational and Mathematical Methods in Data Science (CoMinds) and will focus on Optimization methods for and within Machine Learning.

More information can be found on the workshop homepage.

Summer School (September 11-13 at Zuse Institute Berlin)

Preceding the final conference, a summer school will be held September 11 to 13 focusing both on practical aspects as well as the mathematical background of the topics covered by the Thematic Einstein Semester.

More details can be found on the conference homepage.

Final conference (September 13-15 at Zuse Institute Berlin)

To conclude the activities and achievements of the Thematic Einstein Semester, a final conference will be held September 13 to 15 at the Zuse Institute Berlin that is intended to deepen discussions on selected issues focusing both on theoretical and applied aspects of Optimization and Machine Learning. We hope to explore differences and similarities in existing work and to discover gaps and open questions of interest for future research.

More details can be found on the conference homepage.

Proceedings

Proceedings of the Thematic Einstein Semester will be published by de Gruyter. Every Workshop/Summer School/Conference speaker is entitled to submit a proceedings paper, subject to the following rules:

  • page limit: a length of any article has to be at most 12 pages
  • peer review: by submitting an article you agree to review a different submission
  • deadline: submit until January 15 by sending the PDF to submissions@tes2023.berlin
  • templates: using the de Gruyter LaTeX templates is mandatory. Templates are available upon request.

After acceptance, full LaTeX sources must be submitted as a tar or zip file including figures.