Master theses
Thesis proposals Spring Semester 2023
The following is a list of proposals for master theses offered at the Chair of Risk, Safety and Uncertainty Quantification. To enquire about conducting a project at the Chair, please directly contact the responsible supervisor.
Data-driven regression at the boundary between machine learning and uncertainty quantification
Supervisor: S. Marelli
Most uncertainty quantification (UQ) tools specialize in extracting as much information as possible from very small datasets, minimizing the computational burden of complex engineering models. In contrast, machine learning (ML) methods try to extract sensible structures from a large amount of unstructured available data. Albeit apparently at the opposite ends of the spectrum, the two research fields share many similarities, resulting in significant cross-fertilizations throughout recent years.
This master thesis aims at reviewing the state of the art on data-driven regression from both the UQ and ML literature, to then implement and cross-test their efficiency on standard benchmarks from both fields. In particular, focus will be given to artificial neural networks (ANNs) and model selection strategies from ML, and surrogate models from UQ. If time allows, hybrid extensions to the most promising approaches will be also proposed.
Prerequisistes:
- Good working knowledge of either MATLAB or Python
- Uncertainty quantification in engineering
Additional information
- Group work: No
Advances in recent structural reliability methods
Supervisor: M. Moustapha
The safety of critical civil and engineering systems is of paramount importance to make sure that they can cope with unavoidable environmental and operational uncertainties. Structural reliability analysis provides efficient algorithms to estimate the probability of failure of such systems under various design (e.g., manufacturing tolerance) and operating conditions (e.g., extreme loads).
This master thesis aims at reviewing and benchmarking the recent developments in simulation-based reliability analysis. This includes, for instance, the families of subset simulation and sequential importance sampling, in their latest incarnations such as directional importance sampling on finite-infinite dimensional subset simulation. As a part of the review process, the student will implement the most promising algorithms and compare them with other state of the art techniques, by using benchmark problems and realistic case studies. If time allows, the most efficient algorithm(s) will be implemented (and possibly enhanced through modern tools from the machine-learning literature) and documented in the Uncertainty Quantification software UQLab, to enrich the offer currently available in its structural reliability analysis module.
Prerequisites
- Good working knowledge of Matlab
- Structural Reliability and Risk Analysis
Additional information
- Group work: No
Past theses
The following is a list of past Master theses written at the Chair.
- Tobias Filiberti, Reducing the cost of concrete structures using RBDO, 2022
- Pietro Parisi, Active learning for system rare event estimation, 2021
- Fabian Keller, Surrogate modelling for multiscalethermal simulation of powder-bedadditive manufacturing, 2020
- Riccardo Arrigoni, Uncertainty propagation and sensitivity analysis in hydrology, 2020
- Florian Schmid, A new moment-independent measure for reliability-sensitivity analysis, 2018
- Philippe Wiederkehr, Global Sensitivity Analysis with Dependent Inputs, 2018
- Matteo Berchier, Multi-fidelity surrogate modelling with polynomial chaos expansions, 2016.
- Alessandra Eicher, Bayesian multilevel model calibration of a simplified dam breach model, 2014.