AutumnOLYMPIAD
COLLECTION
RESEARCHED WITH AUTUMN
AR

Adrian Riekert

Mathematician (Dr. rer. nat.); postdoctoral researcher at Mathematics Munster working on the theory of deep-learning optimization; three-time IMO medalist for Germany (2013-2015)
University of Munster · Munster, Germany · academic
LinkedIn
orcid.orgscholar.google.comresearchgate.netdblp.orggenealogy.math.ndsu.nodak.eduuni-muenster.de
Now: Theory of deep-learning optimization: convergence of gradient/Adam methods and operator learning for parametric PDEs
Three-time IMO Medalist For Germany (bronze 2013, Silver 2014, Silver 2015)2024 PhD At University Of Munster On Gradient Methods In Neural Network TrainingWinner Of The 2024 'Theoretical Foundations Of Deep Learning Best Thesis Award' (SPP 2298)Co-author With Arnulf Jentzen On Gradient-method Convergence And Adam-optimizer Theory

Adrian Riekert is a German mathematician and postdoctoral researcher at Mathematics Munster (University of Munster), in the Institute for Analysis and Numerics. He completed his PhD (Dr. rer. nat.) there in 2024 with the dissertation "Mathematical Analysis of Gradient Methods in the Training of Artificial Neural Networks", supervised by Prof. Dr. Arnulf Jentzen, and continued as a postdoc in Jentzen's group. His research centres on deep learning theory, nonconvex optimization, stochastic gradient methods and applied probability. Before his doctorate he studied mathematics at the University of Bonn. As a schoolboy he represented Germany at the International Mathematical Olympiad three times, winning bronze in 2013 and silver in 2014 and 2015.

Dr. Adrian Riekert; PhD 2024 University of Munster (Mathematical Analysis of Gradient Methods in the Training of Artificial Neural Networks, supervisor Arnulf Jentzen); postdoc at Mathematics Munster. IMO Germany team 2013/2014/2015. Winner of the 2024 'Theoretical Foundations of Deep Learning Best Thesis Award' (SPP 2298).

LinkedIn connections157
Details
LocationMunster, Germany
Company siteuni-muenster.de
UniversityUniversity of Munster
Deep learning theoryNonconvex optimizationArtificial neural networksStochastic gradient methodsApplied probabilityNumerical analysisMathematicsMathematical olympiad training

Competition record

Germany · IMO

2013 · Bronze · Rank 1962014 · Silver · Rank 952015 · Silver · Rank 40