I’m a postdoctoral researcher at the Humboldt-University in Berlin. I’m currently part of the department of Knowledge Management in Bioinformatics headed by Ulf Leser and previously was at the machine learning group headed by Marius Kloft who has now shifted to TU Kaiserslautern. My main research interests include machine learning algorithms high-dimensional data in general (texts, gene sequence data, images etc.), Bayesian probabilistic models (in particular topic models), scalable variational inference algorithms, Gaussian processes and Bayesian deep learning models. I received a PhD in natural language processing/machine learning from Leipzig University where my supervisor was Gerhard Heyer. Before that, I completed a Master’s program at Leipzig University and a Bachelor’s program at the University of Cooperative Education in Stuttgart including one term abroad at Staffordshire University’s School of Computing. During my time in Stuttgart, I was also simultaneously employed by IBM Germany where I had the chance to visit IBM’s Almaden research lab in San Jose, California. I also was enrolled at Princeton University as a Visiting Scholar Research Collaborator in Fall 2013. My supervisor there was David Blei (now at Columbia University).


  • Topic modeling
  • Dynamic Bayesian Models
  • Gaussian processes
  • Scalable Baysian Inference


  • PhD in NLP/ML, 2016

    Leipzig University

  • MSc in NLP/ML, 2010

    Leipzig University

  • BSc in Applied Computer Science, 2005

    Stuttgart University of Cooperative Education (BA)


  • Advances in Approximate Bayesian Inference (Workshop at NIPS), since 2015

    Reviewer, Member of Program Committee

  • NIPS conference, 2016, 2017

    Reviewer, Member of Program Committee

  • Open Civil Engineering Journal, since 2016


  • Genomics and Computational Biology, since 2016


  • Digital Humanities Quarterly, since 2016


Recent Publications

More Publications

  • Scalable Generalized Dynamic Topic Models

    Details PDF Code

  • Generalizing and Scaling Up Dynamic Topic Models via Inducing Point Variational Inference

    Details PDF

  • Exploratory Search Through Interactive Visualization of Topic Models

    Details PDF External

  • Exploring Issues in a Networked Public Sphere: Combining Hyperlink Network Analysis and Topic Modeling

    Details PDF External

  • I'm currently not teaching, look out for future lectures or seminars.


  • Office hours: Thursdays, 1 pm - 3 pm
  • +49 (0)30 2093 3167
  • (replace "ä" by "ae")
  • Knowledge Management in Bioinformatics, Humboldt-Universität zu Berlin (Visitors: Rudower Chaussee 25, Bldg. IV, Room 407 (fourth floor), 12489 Berlin; Postal: Unter den Linden 6, 10099 Berlin)