PhD Student · Machine Learning

Hi, I'm Ashutosh Jha

Institute of AI, TU Braunschweig. Advised by Prof. Dr. Michel Besserve.

I develop machine learning methods for understanding complex systems, with a focus on causality, in particular causal and probabilistic representation learning.

Portrait of Ashutosh Jha
Research Interests

Machine Learning for Complex Systems

Probabilistic representation learning: models that capture the structure behind high-dimensional data instead of only predicting it.

Causality

Causal representation learning: recovering the latent variables and mechanisms that generated the observations.

Applications

Sustainable energy systems, decision sciences, econometrics, and generative AI.

News
  • Sep 2026 Started my PhD in Machine Learning at the Institute of AI, TU Braunschweig, supervised by Prof. Dr. Michel Besserve.
  • Aug 2026 Received my M.Sc. in Quantitative Data Science Methods from the University of Tübingen.
  • Aug 2026 Linear Independent Component Analysis via Optimal Transport was accepted at the 9th Workshop on Tractable Probabilistic Modeling at UAI 2026 in Amsterdam.
  • Jul 2026 Preprint of our OT-ICA work is on arXiv.
  • Jun 2026 Submitted my master's thesis, Optimal Transport in linear Independent Component Analysis, at the University of Tübingen and the Max Planck Institute for Intelligent Systems.
Selected Publication

Linear Independent Component Analysis via Optimal Transport

Ashutosh Jha, Michel Besserve, Simon Buchholz

TPM @ UAI 2026 Accepted

All publications →