CV

Click here for the PDF version of my CV.

Education


Employment


Publications

Peer-reviewed papers

Preprints


Teaching

Workshops

Coorganised Workshop Hidden Markov Models for movement, acceleration and other ecological data – an introduction using moveHMM and momentuHMM in R
online, September 2024,
with Roland Langrock and Carlina Feldmann

Practical session for the Workshop Using hidden Markov models to uncover ecological state dynamics from biologging data
Tokyo Japan, March 2024

Tutorials

Practical session Mathematics for economists
Bachelor level, Bielefeld University, summer term 2024

Practical session on hidden Markov models
Master level, Bielefeld University, summer term 2024

Statistical Consulting
Master level, Bielefeld University, winter term 2023/2024

Practical session on generalised linear models (GLMs)
Master level, Bielefeld University, winter term 2023/2024


Grants and Awards


Community involvement

Reviewer


Other skills


Volunteering


Talks and Presentations

Talks

Efficient smoothness selection for nonparametric Markov-switching models via quasi restricted maximum likelihood
24. September 2024, Statistics Seminar, Göttingen, Germany

Exploring hidden dynamics: statistical inference in periodically inhomogeneous hidden Markov models for animal behaviour
4. September 2024, Royal Statistical Society Conference, Brighton, UK

Inference on the state process of periodically inhomogeneous hidden Markov models for animal behaviour
17. July 2024, International Statistical Ecology Conference, Swansea, Wales

Inference on the state process of periodically inhomogeneous hidden Markov models
16. February 2024, Young Researchers Workshop of the Center for Statistics, Bielefeld University
I held a short presentation about our paper Inference on the state process of periodically inhomogeneous hidden Markov models for animal behaviour. We established novel tools to better understand the state process of periodically inhomogeneous hidden Markov models regarding i) the periodically varying unconditional state distribution and ii) the state dwell-time distribution. This was illustrated using the movement track of an elphant from the Ivory Coast. You can download the slides here.

Mitigating consequences of the Markov assumption
24. October 2023, Colloquium of the Center for Statistics, Bielefeld University
I presented the results of my master thesis which focussed on methods for mitigation the undesired consequences of the Markov assumption in hidden Markov models. In particular, I derived properties of the state process of periodically inhomogeneous Markov chains, and extended the model class of hidden semi-Markov models to allow for inhomogeneity in the state dwell-time distributinos. You can download the slides here.

Posters

Mitigating consequences of the Markov assumption
4. December 2023, Doctoral workshop of the Bielefeld Graduate School of Economics and Management (BiGSEM), Bielefeld University
I also presented the results of my master thesis as a poster in the 18. doctoral workshop of the Bielefeld Graduate School of Economics and Management (BiGSEM). You can download the poster here.