Job Opening: Multilevel Architectures and Algorithms in Deep Learning (Ph.D./postdoc)

The SCOOP research group (Scientific Computing and Optimization) at IWR, Heidelberg University is opening a 3-year position. The position is suitable for a Ph.D. student or a postdoc. Applicants are expected to have a background in deep learning and proficiency in programming. Effective verbal and written communication skills are required. Knowledge of numerical optimization, multilevel methods, or adaptive approaches for PDEs are a plus.

The successful candidate will work on the project Multilevel Architectures and Algorithms in Deep Learning led jointly by Roland Herzog and Anton Schiela. This project is part of the DFG Priority Program 2298, a major initiative focusing on the mathematical foundations of deep learning.

The base salary for this position is according to 75% within the TVL E13 pay scale, which amounts to at least 36 600 € p.a. before taxes. A successful postdoctoral applicant will be offered an upgrade to a 100% position (at least 48 800 € p.a. before taxes) from other sources.

Inquiries and applications (including at least a cover letter, CV, and certificates of academic degrees, as a single PDF file) quoting the reference “SPP 2298” should be directed to Roland Herzog. The preferred starting date is January 2022 but a later start is also possible. The position is open until filled.