Kies de Nederlandse taal
Course module: 201900098
Uncertainty quantification and model reduction
Course info
Course module201900098
Credits (ECTS)5
Course typeCourse
Language of instructionEnglish
Contact B. Rosic
Lecturer B. Rosic
Contactperson for the course B. Rosic
Academic year2021
Starting block
Application procedureYou apply via OSIRIS Student
Registration using OSIRISYes
After this course the student is able:
  • to remember the concept of uncertainties, their modelling and quantification
  • to understand the difference between aleatoric and epistemic uncertainties, forms of their modelling and reduction
  • to select and implement procedures for efficient propagation of uncertainties in linear and nonlinear systems
  • to reduce the dimension of the uncertain outcome
  • to analyze and validate different quantification techniques
  • to synthesize results of the analysis into a meaningful conclusion for engineers, decision makers and field scientists.
The data driven modelling has for a goal to transform large amounts of data with the help of machine/artificial learning techniques to the practically usable predictive models. However, the resulting predictive analytics is known to be sensitive to the size of training data sets, the design of machine architecture, the type of chosen learning method, modelling errors etc. On the other hand, the classical physics is focusing on more detailed modelling of processes governed by well-understood laws of physics that are in nature deterministic, and already understood to great maturity. In most of occasions, however, these cannot be fully specified due to lack of experimental data or knowledge on the parameter values, boundary conditions and excitations. To address these modelling issues, the main goal of the course is to give students a general overview on how aleatoric (natural variability) and epistemic (lack of knowledge) uncertainties can be described and incorporated to the modelling process, and further used to obtain more reliable and robust model predictions. Furthermore, the course will also address sensitivity analysis and model reduction as additional parametric dependency on uncertainties gives rise to the system dimension by several order of magnitude. More importantly, in this course students will get practical knowledge in quantifying and analyzing uncertainties in real engineering systems by working on different engineering applications ranging from dynamical systems up to the continuum mechanics.
Assumed previous knowledge
Mandatory: Students need to have a solid background in multivariate calculus and linear algebra (e.g. Linear Algebra, 201500292), Probabilty Theory and Statistics (201400221), Statistics and Probability (191506103), or similar. In addition the basic programming skills in Python or Matlab are required.
Recommended: Students with a theoretical focus, will find skills obtained from the MSc math courses scientific computing, information theory or complex networks helpful for this.
Participating study
Master Mechanical Engineering
Participating study
Master Systems and Control
Participating study
Master Electrical Engineering
Participating study
Master Civil Engineering and Management
Participating study
Master Applied Mathematics
Participating study
Master Computer Science
Required materials
Xiu, D. (2010) Numerical methods for stochastic computations - A spectral method approach, Princeton University press. ISBN 9781400835348
Le Maitre, Oliver and Knio, Omar M. (2010) Spectral methods for uncertainty quantification with applications to computational fluid dynamics. ISBN 978-90-481-3520-2, Springer Netherlands
Course material
Lecture slides and selection of papers
Recommended materials
Instructional modes
Presence dutyYes

Presence dutyYes

Assignment(s) and oral exam

Kies de Nederlandse taal