PhD Position in Hierarchical Bayesian Inference using Stochastic Emulators

The Chair of Risk, Safety & Uncertainty Quantification (RSUQ) at ETH Zurich develops cutting-edge methodologies in the field of uncertainty quantification (UQ) for engineering systems. Our research covers surrogate modeling, reliability analysis, sensitivity analysis, optimization under uncertainty, and Bayesian calibration. We are known for developing the UQLab software framework for uncertainty quantification, which is widely used in academia and industry.

This PhD position is part of the ORACLES project ("Optimization, Reliability And CaLibration using Emulators of Stochastic computational models"), funded by the Swiss National Science Foundation (SNSF). The project aims to significantly advance the state-of-the-art in uncertainty quantification (UQ) by developing and applying novel stochastic emulators. A key focus is using these advanced emulators to tackle complex, UQ problems previously intractable with deterministic simulators, such as high-dimensional reliability-based design optimization and hierarchical Bayesian inversion. This specific PhD position focuses on the challenges within hierarchical Bayesian inference.

As the successful candidate, your research will focus on developing and applying advanced computational methods for Bayesian inference and model calibration. Your main tasks will include:

• Investigating and developing novel computational strategies to enhance the efficiency and scalability of Bayesian methods for complex models
• Exploring the use of advanced surrogate modeling techniques (e.g., stochastic emulators) to accelerate likelihood computations and posterior exploration in Bayesian frameworks
• Addressing challenges related to high-dimensional parameter spaces and complex, potentially high-dimensional model outputs in Bayesian analysis
• Developing methods to improve the accuracy and robustness of parameter estimation and uncertainty quantification using Bayesian techniques
• Applying the developed methods to calibrate and validate computational models against experimental data in relevant engineering contexts
• Disseminating research findings through publications in leading peer-reviewed journals and presentations at international conferences

We are looking for a highly motivated candidate with:

• A Master's degree in Computational Science/Engineering, Applied Mathematics, Mechanical Engineering, Civil Engineering, or a related field
• A strong background and keen interest in uncertainty quantification, structural reliability, optimization, Bayesian inference, surrogate modeling/emulation, and/or machine learning
• Excellent programming skills (preferably Matlab, or Python)
• Strong analytical and problem-solving abilities
• Excellent communication and scientific writing skills, and fluency in English (both written and spoken)
• Enthusiasm for pursuing cutting-edge research within an international, multicultural collaborative team

Within the Department of Civil, Environmental and Geomatic Engineering, the Chair of Risk, Safety and Uncertainty Quantification offers an exciting opportunity to join a small, highly international research group of around 12 members. We foster a collaborative, supportive atmosphere where open scientific exchange and mutual respect are key. Our working culture is flexible and results-oriented, offering you the freedom to manage your schedule in a way that supports both productivity and personal well-being. Regular group discussions, interdisciplinary collaborations, and a shared passion for research make this an inspiring environment for pursuing a PhD.