Research directions

• Development of rigorous data analysis methods and their application in quantitative biology (methods: probabilistic latent feature models, neural networks, Random Matrix Theory, dimensionality reduction, …).
• Inference methods in the context of risk-minimizing collective betting strategies (methods: Bayesian inference, optimisation, Information Geometry, …).

Completed projects

• Playing it safe: information constrains collective betting strategies. Preprint with Vijay Balasubramanian.
  How should you place bets when you do not know the odds of the game? We use Bayesian inference theory to show that risk can be reduced by assuming a simpler probabilistic model. We define what simple means using Information Geometry and derive our result for the class of exponential family probability distributions.
• Hidden data structures and statistics of natural images. PRX Life with Ilya Nemenman.
  Whether sparse, with clusters, or overlapping clusters – large real-world data can contain many types of hidden patterns. Is there a simple statistical model to generate and understand such data? We construct such a model and use it to show what we can learn about data just from correlations and apply our method to natural images.
• Random Matrix model with latent features and noise. PRE article with Ilya Nemenman.
  We use Principal Component Analysis all the time, but when is it actually useful? How does data look if it comes from a model with a handful of linear latent features? In our work we give the answers.
• Biomechanical generation of behaviour in a living network. eLife article with Mirna Kramar, Michael Wilczek and Karen Alim.
  We study the biomechanical generation of behaviour in the single-cell organism Physarum Polycephalum by analysis of bright-field microscopy movies.
• Curious exploration strategies in complex environments via reinforcement learning. See the IEEE proceedings with Menachem Stern, Clélia de Mulatier and Vijay Balasubramanian for first results.

Gravity, String Theory, and Automorphic Functions

Prior to entering research in Biophysics at the end of 2016, I was working in the field of theoretical High Energy Physics. More precisely, I was working on research problems related to Gravity, String Theory and the mathematical theory of Automorphic Functions (see my publications). I have also co-authored a book on the topic of “Eisenstein Series and Automorphic Representations’’, providing a survey of this vast subject with a focus on applications in String Theory. The book was published by Cambridge University Press in July 2018.