Uncertainty Quantification and Predictive Modeling

Predictive modelling and uncertainty quantification are more than just another research direction relevant to science and engineering. They constitute a different way of thinking that impacts practically all aspects of scientific and engineering analysis and design. Rather than deriving deterministic answers to complex problems, probabilistic distributions of outcomes are obtained that account for our incomplete and often inaccurate information about the problems of interest.

The mission of the Uncertainty Quantification and Predictive Modeling working group is to develop computational, mathematical and statistical methodologies towards predictive modelling applied to a broad range of applications. The developed methodologies explore synergistic advances in mathematical and statistical sciences, machine learning and computational science. Particular fundamental problems of interest include methods for addressing the curse of stochastic dimensionality, stochastic coarse graining in multiscale/multiphysics simulations, data-driven science, modeling of rare events, solution of multiscale inverse problems, and other. To capitalize at TUM’s strengths and expertise in materials and Continuum Mechanics, one of the main themes of interest is in predictive materials science but other applications in engineering and sciences are also considered (reliability and risk structural analysis, fluid-structure interaction, transport processes in random media, etc.).

The TUM-IAS Focus Group on Uncertainty Quantification and Predictive Modelling offers a unique opportunity to bring together diverse communities within TUM that have common interests in predictability of diverse complex systems requiring similar underpinning mathematical, statistical and algorithmic developments.

Prof. Nicholas Zabaras is the Chair of Uncertainty Quantification at the University of Warwick and the founding Director of the Warwick Centre for Predictive Modelling (WCPM). He is an international leader in the field of uncertainty quantification in particular as it applies to multiscale systems. He has worked extensively on Bayesian approaches for surrogate model development, stochastic model reduction and inverse problems. As a Hans Fischer Senior Fellow at the TUM-IAS, Prof. Zabaras has joined the Uncertainty Quantification and Predictive Modeling Group. Prof. Phaedon-Stelios Koutsourelakis Continuum Mechanics) acts as a host at TUM for this group.

Isabell Franck (Continuum Mechanics) also worked as a doctoral candidate in this group.

TUM-IAS funded doctoral candidate:
Markus Schöberl, Continuum Mechanics

Publications by the Focus Group


  • M. Schöberl, N. Zabaras, and P. S. Koutsourelakis: Bayesian Coarse-Graining. 2017 more…
  • M. Schöberl, N. Zabaras, and P. S. Koutsourelakis: Bayesian coarse-graining in atomistic simulations: adaptive identification of the dimensionality and salient features. 2017 more…
  • Schöberl, Markus; Zabaras, Nicholas; Koutsourelakis, Phaedon-Stelios: Predictive coarse-graining. Journal of Computational Physics 333, 2017, 49-77 more…


  • Koutsourelakis, P.S.; Zabaras, N.; Girolami, M.: Special Issue: Big data and predictive computational modeling. Journal of Computational Physics 321, 2016, 1252-1254 more…