A post doc position in theory of machine learning is available.
The post doc is under the supervision of Professor Kasper Green Larsen, Aarhus University, Denmark. The focus of the research project is on understanding and improving the performance of classic learning algorithms, in particular Boosting and Bagging, both in terms of speed and generalization capabilities. The project also allows pursuing other ideas and directions within theory of machine learning if the candidate has preferences or ideas from prior work.

The research is carried out in collaboration with Kasper as well as Ph.D. students and post docs in the group. The candidate is expected to have a strong theoretical background and to be able to carry out proofs independently or with minor guidance. It is not expected that the candidate will do implementations, however, if the candidate wishes to do so, and it is natural in the project, then implementation work is most welcome.
In addition to research, there may be minor teaching activities, such as presenting at seminars in the research group or guiding Ph.D. students in joint projects.
Candidates must have expertise in theory of machine learning and/or theoretical computer science. Publications at top venues such as COLT/NeurIPS/ICML or STOC/FOCS/SODA are much appreciated.