Picture of Dr. Martin Eigel

MATHEON Logo

Dr. Martin Eigel

Short CV Publications Teaching

Research interests

Association to the Mathematical Topic "Numerical Methods for PDEs with Stochastic Data".

  • Statistical and Deep Learning
  • Adaptive functional approximations for stochastic PDE
  • Low-rank tensor methods for stochastic PDE
  • Statistical inverse problems
  • Topology and shape optimisation under uncertainties
  • FEM a posteriori error estimators
  • Spatial models in computational biology

Contact details

E-mail Martin.Eigel-please remove this text-@wias-berlin.de
Phone +49 (0) 30 20372 413
Fax +49 (0) 30 20372 412

Current projects

  • Machine Learning in tensor formats: Analysis and development of statistical, deep, reinforcement learning by means of adapted tensor representations.
  • Adaptive numerical methods for stochastic PDE: Adaptive spectral methods for problems with stochastic data.
  • Low-rank methods for Stochastic FEM: Efficient adaptive low-rank tensor solvers for SGFEM discretisations of stochastic PDE.
  • Functional Bayesian inversion: Inverse problems with stochastically perturbed measurements based on low-rank tensor approximations.
  • Multi-scale failure analysis with polymorphic uncertainties for optimal design of rotor blades: DFG SPP1886 sub-project 4.
    Head of project together with D. Hömberg and J. Petryna.

Short CV

Since March 2013Researcher in group Hömberg, WIAS, Berlin
November 2010 - February 2013PostDoc in group Peterseim, Matheon project C33, Humboldt Universität, Berlin
July 2008 - November 2010PostDoc in group Carstensen, Humboldt Universität, Berlin
2008Ph.D., University of Warwick, United Kingdom
June 2003Diploma, Universität Heidelberg

Publications

Articles

Proceedings

Thesis

  • M. Eigel, An Adaptive Meshfree Method for Reaction-Diffusion Processes on Complex Domains,
    University of Warwick, Ph.D. Thesis, 2008
  • M. Eigel, Numerische Simulation von Transportvorgängen in der Zelle,
    Universität Heidelberg, Diploma Thesis, 2003

Teaching

Uncertainty Quantification and Statistical Learning WS 2018/19 (with R. Schneider, TUB)

Realisierungen von Zufallsfeldern
(30.10.) Die Vorlesung findet diese Woche entgegen der Information im Vorlesungsverzeichnis statt!
(19.10.) Info Raumänderung: dienstags ab jetzt in E-N 193!
(22.12) Hier noch der letzte kleine Teil zur Vorlesung über a posteriori Fehlerschätzer. Alles Gute für 2019!

Referenzen

Uncertainty Quantification and Parametric PDEs
Bayesian Inversion
Statistical Learning, Hierarchical Tensors and Neural Nets

Uncertainty Quantification and Tensor Approximation WS 2016/17 (with R. Schneider, TUB)