photo Philippe von Wurstemberger
ETH Zurich


Address:
Philippe von Wurstemberger
Seminar for Applied Mathematics
Department of Mathematics
ETH Zurich
Rämistrasse 101
8092 Zürich
Switzerland

Office: Room HG E 62.1
Phone: +41 44 632 0401
Fax: +41 44 632 1104

E-mail: philippe.vonwurstemberger (at) sam.math.ethz.ch
Homepage: vonwurstemberger.ch

Links: [Profile on ResearchGate] [Profile on GoogleScholar]

Research interests

  • Approximation methods for high dimensional PDEs
  • Mathematical analysis of deep learning (Approximation capacities of artificial neural networks, stochastic optimization)
  • Reinforcement learning

Education

  • since 10/2018:      PhD student in the research Group of Prof. Arnulf Jentzen, ETH Zurich, Switzerland
  • 10/2018:               Master of Science ETH in Mathematics, ETH Zurich, Switzerland
  • Fall 2015:             Exchange Semester in Princeton, Princeton University, USA
  • 03/2015:               Bachelor of Science in Mathematics, ETH Zurich, Switzerland

Preprints

  • Hutzenthaler, M., Jentzen, A., von Wurstemberger, P., Overcoming the curse of dimensionality in the approximative pricing of financial derivatives with default risks. [arXiv] (2019), 71 pages.
  • Hutzenthaler, M., Jentzen, A., Kruse, T., Nguyen, T. A., von Wurstemberger, P., Overcoming the curse of dimensionality in the numerical approximation of semilinear parabolic partial differential equations. [arXiv] (2018), 27 pages.
  • Grohs, P., Hornung, F., Jentzen, A., von Wurstemberger, P., A proof that artificial neural networks overcome the curse of dimensionality in the numerical approximation of Black-Scholes partial differential equations. [arXiv] (2018), 124 pages.
  • Jentzen, A., von Wurstemberger, P., Lower error bounds for the stochastic gradient descent optimization algorithm: Sharp convergence rates for slowly and fast decaying learning rates. [arXiv] (2018), 42 pages.
  • Jentzen, A., Kuckuck, B., Neufeld, A., von Wurstemberger, P., Strong error analysis for stochastic gradient descent optimization algorithms. [arXiv] (2018), 75 pages. Revision requested from IMA J. Num. Anal.