Home | deutsch  |  Impressum  |  Data Protection  |  Sitemap  |  Intranet  |  KIT
Research Group 3: Scientific Computing

Secretariat
Kollegiengebäude Mathematik (20.30)
Room 3.039

Address
Hausadresse:
Zimmer 3.039
Englerstr. 2
Kollegiengebäude Mathematik (20.30)
76131 Karlsruhe

Postadresse:
Karlsruher Institut für Technologie (KIT)
Fakultät für Mathematik
Institut für Angewandte und Numerische Mathematik
Arbeitsgruppe 3: Wissenschaftliches Rechnen
Englerstr. 2
Kollegiengebäude Mathematik (20.30)
D-76131 Karlsruhe

Office hours:
Mon-Thu 9-12 Uhr

Tel.: +49 721 608 42062

Fax.: +49 721 608 43197

Photo of Tudor Udrescu Dr. Tudor Udrescu

Office hour for students:
Room: Kollegiengebäude Mathematik (20.30)
Email: tudor.udrescu@kit.edu




Current List of Courses
Semester Titel Typ
Summer Semester 2013 Lecture
Lecture
Winter Semester 2012/13 Lecture
Seminar
Winter Semester 2011/12 Seminar
Winter Semester 2010/11 Seminar


Research

My focus lies on the construction of numerical methods for stochastic reaction networks modeling the dynamics of biochemical processes.

A stochastic model uses randomness to describe hugely complex biophysical processes like regulation of gene expression or intracellular viral kinetics, with the mathematical formulation being provided by a Markov jump process. The corresponding probability distribution is the main object of interest in stochastic reaction kinetics and is obtained as the solution of the Chemical Master Equation (CME). However, the task of solving this equation is not trivial due to the curse of dimensionality: the computational effort grows exponentially with the number of species involved in the reaction network.

I work on mitigating the effects of the curse of dimensionality and make the CME and other high-dimensional problems computationally tractable.

Research interests

  • Stochastic computation in biology and mathematical finance
  • High-dimensional approximation using adaptive wavelet compression
  • Qualitative analysis of rare events in metastable systems
  • Complexity reduction using hybrid stochastic-deterministic models
  • Effective solvers (projection and multigrid algorithms)
  • Parallelization of computational methods

More information about my research area can be found here.


PhD Thesis

Tudor Udrescu. Numerical Methods for the Chemical Master Equation, PhD Thesis, online in Elektronischen Volltext-Archiv (EVA), Karlsruhe Institute of Technology (KIT), 2012.


Publications

Articles

  • Gheorghe Juncu, Aurelian Nicola, Constantin Popa, Tudor Udrescu. Nonlinear multigrid methods for solving Richards’ equation in two space dimensions. Carpathian Journal of Mathematics, 25(1), 82-91 (2009).

In proceedings

  • Constantin Popa and Tudor Udrescu. A fast approximation for discrete Laplacian, Proceedings of the 4th Workshop on Mathematical Modelling of Environmental and Life Sciences Problems, Constanta 2005, Publishing House of the Romanian Academy, 181-190 (2006).


Miscellaneous

Eugen Petac and Tudor Udrescu. Fundamente Java (Java Fundamentals, in Romanian), MatrixRom Publishing House, Bucharest 2005, 500 pag., ISBN 973-685-944-4

Selected Talks