Jan Bölts

I am a PhD student in the mackelab at University of Tübingen working on the intersection of machine learning and neuroscience. I have a background in Cognitive Science (BSc) and Computational Neuroscience (MSc) and discovered my passion for Bayesian statistics and machine learning during my masters. Since then I am developing probabilistic machine learning tools and applying them to problems in computational neuroscience.

My research focus lies on simulation-based inference (SBI), a method to perform Bayesian inference in simulation-based models like they often occur in astrophysics, economics, biology, neuroscience and many other fields. I am one of the core developers of sbi, a Python toolbox for SBI (link to paper). At the moment I am particularly interested in hierarchical inference, model misspecification and model comparison in SBI. Please contact me for questions and dicussions.

Interests

• Bayesian inference
• Simulation-based inference
• Probabilistic programming
• Bayesian deep learning

Education

• PhD in Machine Learning, present - University of Tübingen
• MSc in Computational Neuroscience, 2018 - Technical University Berlin
• BSc in Cognitive Science, 2015 - University of Osnabrück