PhD positions on machine learning in relativistic optics

Webseite Centre for Advanced Laser Applications, LMU München

Introduction

The invention of Chirped-Pulse Amplification (Nobel prize Strickland & Mourou, 2018) has led to a dramatic performance increase in modern laser systems. These lasers can reach peak powers of several petawatt, which momentarily rivals the entire electrical output of our planet. When such laser pulses are focused, the resulting intensity is high enough to immediately transform all matter in vicinity to the focus into a relativistic plasma. The interaction between laser and plasma leads to numerous interesting physical phenomena (laboratory astrophysics, QED studies, etc.) and applications such as ultra-compact laser-driven accelerators and x-ray sources.

However, due to the extreme conditions and low repetition rate of high-power laser systems, it is difficult to diagnose and optimize experiments. Our group is exploring a variety of machine learning methods to improve the performance of such laser-plasma experiments. This includes the development of novel Bayesian optimization methods (Irshad et al. 2021, recently included in the BoTorch package by Facebook/Meta), expectation-maximization for tomographic reconstruction (A.Döpp et al. 2018) and various neural network architectures such as CNNs, ResNets or U-Nets for advanced diagnostics (to be published).

With this call we are looking for candidates to join our team to work on applications of machine learning in relativistic plasma optics. The project is based at the Centre for Advanced Laser Applications (CALA) near Munich, affiliated to the chair of Prof. Ferenc Krausz (LMU Munich University / Max-Planck Institute for Quantum Optics). As a safe, clean, and cosmopolitan city close to the German alps, Munich is ranked frequently among the most livable cities in the world. The employment contract will be with LMU Munich, which is frequently ranked among the best universities in Germany and one of the top universities in physical sciences worldwide. PULSE, the petawatt physics laboratory at CALA, hosts one of the most powerful laser systems in the world, the ATLAS-3000 multi-petawatt laser, and regular access to the laser within the project is assured. Research will be conducted in close collaboration with researchers at University of Oxford in the UK, with the opportunity for research visits.

Profile

To expand our team we are looking for applications by talented and motivated PhD candidates with the following qualifications:

  • Master’s degree in physics, mathematics, computer science or engineering
  • Interest, and ideally experience, in experimental laser physics and optical physics
  • Experience in programming (mainly Python) and machine learning is beneficial
  • Fluent in English

Qualified applicants will be given an assessment test, followed by an interview round. The successful applicants will receive 3-year contracts (salary calculated according to the Germany’s public TVL system), with employment ideally starting in autumn of 2022.

Application

Questions and applications (including a short letter of motivation, a CV, university certificates with grades, as well as names and email addresses of two potential references) should be addressed to the project coordinator, Dr. Andreas Döpp. Submission deadline is July 31, 2022.

Please send all messages to atlas@physik.uni-muenchen.de

Further reading

  • F. Irshad, S. Karsch and A. Döpp, Expected hypervolume improvement for simultaneous multi-objective and multi-fidelity optimization, arXiv 2112.13901 (2021)
  • A. Döpp, L. Hehn, et al., Quick x-ray microtomography using a laser-driven betatron source, Optica Vol. 5, Issue 2, pp. 199-203 (2018)
  • J. Wenz, A. Döpp, et al, Dual-energy electron beams from a compact laser-driven accelerator. Nature Photonics, Vol. 13, Pages 263–269 (2019)
  • J. Götzfried, A. Döpp et al., Physics of High-Charge Electron Beams in Laser-Plasma Wakefields, Phys. Rev. X 10, 041015 (2020)

Um dich für diesen Job zu bewerben, besuche bitte www.pulse.physik.uni-muenchen.de.