Back to home

Curriculum vitae

Marek Jankola
Marek.Jankola@ista.ac.at

Education

PhD in computer science · 2025–present

Institute of Science and Technology Austria (ISTA)
Advised by Marco Mondelli and Francesco Locatello.

MSc in theoretical physics · 2023–2025

Faculty of Mathematics and Physics, Charles University, Czech Republic
Graduated summa cum laude. Master’s thesis research at EPFL’s Statistical Physics of Computation laboratory, supervised by Lenka Zdeborová.

BSc in general physics · 2019–2023

Faculty of Mathematics and Physics, Charles University, Czech Republic
Graduated summa cum laude.

Research

Learning dynamics, high-dimensional statistic, identifiability and causal discovery. During my first year at ISTA, I studied identifiability dynamics in online ICA performed by shallow autoencoder and its applications to causal discovery, performative classification in high-dimensional Gaussian mixtures, and theoretically principled risk curves in statistical models for genetic data.

Optimizing initialization in graph dynamics. In my master’s thesis at EPFL, I developed an efficient algorithm based on the backtracking dynamical cavity method and belief propagation reinforcement to find initial conditions leading to a desired outcome in dynamical graphical models.

Entropy production in periodically driven systems. In my bachelor’s thesis, I derived and proved a new bound on entropy production in periodically driven systems within stochastic thermodynamics.

Teaching and volunteering

Skills

Python, MATLAB, C++, LaTeX and Mathematica.
English (C1); beginner Spanish.