Profile
Transcrição
Profile
Michael Griebel Date of birth: 13 Jan 1960 Academic career 1985 1985 - 1990 1989 1990 - 1993 1993 1994 - 1995 1995 1995 - 1995 2000 2009 Since 1996 Since 2003 Since 2010 Diploma in Computer Science, TU Munich (advisor: F. L. Bauer) Research Assistant, Institute for Informatics, TU Munich PhD in Computer Science, TU Munich (advisor: C. Zenger) Assistant Professor (C1), Institute for Informatics, TU Munich Habilitation in Computer Science, TU Munich (advisor: C. Zenger) Assistant Professor (C2, Wissenschaftlicher Oberassistent), Institute for Informatics, TU Munich Venia Legendi in Computer Science, TU Munich Priv.-Doz., Institute for Informatics, TU Munich Visiting Professor, University of California, San Diego, CA, USA Visiting Professor, Paris Diderot University (Paris 7), France Professor (C4) of Scientific Computing and Numerical Simulation, University of Bonn Director, Institute for Numerical Simulation, University of Bonn Director, Fraunhofer-Institute for Algorithms and Scientific Computing (SCAI), Bonn Honours 2004 2007 2017 Fellow, Institute for Pure and Applied Mathematics, University of California, Los Angeles, CA, USA International Fellow, Australian Research Council (ARCIF), University of New South Wales, Sydney, NSW, Australia John von Neumann Visiting Professorship, TU Munich Offers 1995 1998 1998 2002 Invited Lectures C4 professorship in numerical mathematics, Düsseldorf C4 professorship in simulation of large systems, Stuttgart Research position at the Lawrence Livermore National Laboratory, California, USA C4 professorship in technomathematics, TU Kaiserslautern joint with leading position at the Fraunhofer Institute for Industrial Mathematics (Fraunhofer-Institut für Techno- und Wirtschaftsmathematik, ITWM) 2006 2006 2006 2006 2007 2007 2008 2008 2009 2010 2012 2016 International Congress on the Applications of Mathematics, Santiago de Chile, Chile World Congress on Computational Mechanics, Los Angeles, CA, USA International Congress of Mathematicians, Madrid, Spain Numerical Methods in Finance. An AMaMeF Conference, INRIA, Rocquencourt, France ENUMATH, Graz, Austria European Postgraduate Fluid Dynamics Conference, Birmingham, England, UK IUTAM, Symposium on Modeling Nanomaterials and Nanosystems, Aalborg, Denmark Summer School on Nanotechnology and Mathematics, Santiago de Compostela, Spain Algorithms and Complexity for Continuous Problems, Dagstuhl, Wadern Zürich Summer School, Sparse Tensor Discretizations of High-Dimensional Problems, Switzerland ESF-JSPS Frontier Science Conference for Young Researchers: Mathematics for Innovation Large and Complex Systems, Tokyo, Japan SIAM Conference on Uncertainty Quantification, Lausanne, Switzerland Research profile A main focus of current research is the approximation of high-dimensional functions and the solution of partial differential equations in high dimensions. Our central tool in this respect is the dimension-adaptive sparse grid method and its generalizations. Another focus is the development of efficient numerical methods for the treatment of two-phase-flow problems and for the simulation of materials on the atomistic and the continuous scale with special emphasis on nano-technology. Finally, adaptive discretizations for partial differential equations, multi-level solvers of the arising linear systems and their parallelization have been investigated. A main focus of future research will be numerical data analysis. This involves high dimensional regression, density estimation and classification problems in high-dimensional spaces. The emphasis will be on the development of new numerical techniques and on proposing stable algorithms as well as on establishing theoretical results. The research will be driven by real data applications in econometrics and finance and by data-oriented problems from material science. Since data often stem from a low-dimensional manifold embedded in ambient space, we aim at exploiting this structure by applying adaptive sparse grid algorithms for the h-version, the p-version and the kernel-based versions of approximation. Moreover, we will develop nonlinear methods beyond the Hilbert space setting. To this end, we will replace the conventional Euclidean cost function by more appropriate distance measures such as Banach space norms or Bregman divergences. Here, we again aim at deriving both, a theoretical foundation and stability conditions with error estimates for practical purposes. Finally, we plan to study the relation of our nonlinear methods to deep neural networks. Editorships • Numerische Mathematik (Managing Editor) • Springer Lecture Notes in Computational Science and Engineering • Springer Texts in Computational Science and Engineering Research Area J A main focus of research is the approximation of high-dimensional functions and the solution of partial differential equations in high dimensions. Also, data analysis problems are of interest here. Our central tool in this respect is the dimension-adaptive sparse grid method and its generalizations. Another focus is the development of efficient numerical methods for the treatment of two-phase-flow problems and for the simulation of materials on the atomistic and the continuous scale with special emphasis on nano-technology. Finally, adaptive discretizations for partial differential equations, multi-level solvers of the arising linear systems and their parallelization are under investigation. Selected PhD students Frank Kiefer (2001): “Multiskalen-Verfahren für Konvektions-Diffusions Probleme”, now Programme Director, DFG Jochen Garcke (2004): “Maschinelles Lernen durch Funktionsrekonstruktion mit verallgemeinerten dünnen Gittern”, now Professor, University of Bonn Jan Hamaekers (2009): “Tensor Product Multiscale Many-Particle Spaces with Finite-Order Weights for the Electronic Schödinger Equation”, now Head of Department ”Virtual Material Design”, Fraunhofer SCAI Habilitations Gerhard Zumbusch (2001), now Professor, University of Jena Thomas Gerstner (2007), now Professor, University of Frankfurt Marc Alexander Schweitzer (2008), now Professor, University of Bonn Christian Rieger (2016) Selected publications [BG04] [BG09] [GG03] [GH10a] [GH10b] [GK09] [GO95] [GO07] [Gri94] [Gri06] [GW06] B UNGARTZ, Hans-Joachim ; G RIEBEL, Michael: Sparse grids. In: Acta Numer. 13 (2004), S. 147–269. – ISSN 0962–4929 B RAUN, Jürgen ; G RIEBEL, Michael: On a constructive proof of Kolmogorov’s superposition theorem. In: Constr. Approx. 30 (2009), Nr. 3, S. 653–675. – ISSN 0176–4276 G ERSTNER, T. ; G RIEBEL, M.: Dimension-adaptive tensor-product quadrature. In: Computing 71 (2003), Nr. 1, S. 65–87. – ISSN 0010–485X G RIEBEL, M. ; H AMAEKERS, J.: Tensor Product Multiscale Many-Particle Spaces with Finite-Order Weights for the Electronic Schrödinger Equation. In: Zeitschrift für Physikalische Chemie 224 (2010), S. 527–543 G RIEBEL, Michael ; H OLTZ, Markus: Dimension-wise integration of high-dimensional functions with applications to finance. In: J. Complexity 26 (2010), Nr. 5, S. 455–489. – ISSN 0885–064X G RIEBEL, M. ; K NAPEK, S.: Optimized general sparse grid approximation spaces for operator equations. In: Math. Comp. 78 (2009), Nr. 268, S. 2223–2257. – ISSN 0025–5718 G RIEBEL, M. ; O SWALD, P.: On the abstract theory of additive and multiplicative Schwarz algorithms. In: Numer. Math. 70 (1995), S. 163–180. – Also as SFB-Report 342/6/93 A, Institut für Informatik, Technische Universität München, 1993 G RIEBEL, M. ; O ELTZ, D.: A sparse grid space-time discretization scheme for parabolic problems. In: Computing 81 (2007), Nr. 1, S. 1–34. – ISSN 0010–485X G RIEBEL, Michael: Multilevel algorithms considered as iterative methods on semidefinite systems. In: SIAM J. Sci. Comput. 15 (1994), Nr. 3, S. 547–565. – ISSN 1064–8275. – Iterative methods in numerical linear algebra (Copper Mountain Resort, CO, 1992) G RIEBEL, Michael: Sparse grids and related approximation schemes for higher dimensional problems. In: Foundations of computational mathematics, Santander 2005 Bd. 331. Cambridge Univ. Press, Cambridge, 2006, S. 106–161 G RIEBEL, M. ; W OZNIAKOWSKI, H.: On the optimal convergence rate of universal and non-universal algorithms for multivariate integration and approximation. In: Mathematics of Computation 75(255) (2006), S. 1259–1286