Research associate / PhD position in model order reduction and Bayesian inverse problems

Universität/Firma :Universität Potsdam
Fachgebiet :Mathematics
Kontakt :hanlie@uni-potsdam.de
Webadresse :https://www.sfb1294.de/fileadmin/user_upload/Open_positions/A07b_2ndround.pdf

Ausschreibungstext:

The DFG-funded Collaborative Research Center SFB 1294 “Data Assimilation – The Seamless Integration of Data and Models”, hosted at the University of Potsdam jointly with its partner institutions HU Berlin, TU Berlin, WIAS Berlin and GFZ Potsdam, invites applications for a doctoral researcher position (75% of full-time employee position with salary grade TV L - E13) within Project A07: “Model order reduction for Bayesian inference”.

The candidate will work at the Institute of Mathematics at the University of Potsdam under the supervision of Prof. H. C. Lie.  The candidate will closely collaborate with the group of Prof. M. Freitag (Institute of Mathematics, University of Potsdam).

Within Project A07, the doctoral researcher will develop the mathematical theory of Bayesian inference with dimension reduction, in the context of inverse problems. They will analyse the effect of low-rank approximations and projection-based dimension reduction on the posterior measure for infinite-dimensional parameter spaces. They will also develop and analyse algorithms for Bayesian inference using model order reduction. The broader aim of this project is to develop theoretically validated and computationally efficient methods for Bayesian inference that involve high-dimensional spaces and/or expensive forward models. Three relevant references are:

1. T. Cui, Y. Marzouk, and K. Willcox. "Scalable posterior approximations for large-scale Bayesian inverse problems via likelihood-informed parameter and state reduction". Journal of Computational Physics, 315:363 – 387, 2016.

2. A. K. Saibaba, J. Chung, and K. Petroske. "Efficient Krylov subspace methods for uncertainty quantification in large Bayesian linear inverse problems". Numerical Linear Algebra and Applications, page e2325, 2020.

3. A. Spantini, A. Solonen, T. Cui, J. Martin, L. Tenorio, and Y. Marzouk. Öptimal low-rank approximations of Bayesian linear inverse problems". SIAM Journal of Scientific Computing, 37(6):A2451–A2487, 2015.

More information about Project A07 is available at https://www.sfb1294.de/research/research-area-a.

The ideal candidate has mastered measure-theoretic probability, functional analysis, and linear algebra. They have experience in numerical analysis and Markov Chain Monte Carlo methods, and a strong interest in rigorous mathematical analysis and its application to numerical methods for inverse problems. The candidate can provide convincing evidence of these qualifications by coursework, research projects, and/or a master’s thesis. In addition, the candidate has experience with scientific computing in Matlab or Python. The candidate has a strong ability to work effectively, both in collaboration with others and independently. The candidate must be able to communicate effectively in both written and spoken English.

The SFB 1294 provides an excellent research infrastructure including a large interdisciplinary network of researchers and its own graduate school, as well as funding opportunities for conference visits, summer schools, and hosting international experts.

The SFB 1294 seeks to promote diversity in research, and encourages qualified applicants of any gender and from any background to apply.

Applications to the SFB should be submitted via https://www.geo-x.net/sfb-1294/ and should include the following in a single PDF file:

    A statement of research interests and motivation;
    A full CV;
    The names, e-mail addresses and/or reference letters of at least two referees;
    Academic transcripts;
    A link to an electronic version of your Master/Diploma thesis; and
    A list of publications/talks/presentations.

In your application, please indicate that you are applying for the doctoral researcher position in project A07b, and explain your motivation for applying to this position.


Diese Anzeige wurde am 03/09/2021 um 11:09 übermittelt und verfällt am 01/10/2021.

Zurück zur Jobbörse.