Andreas Tersenov
PHD STUDENT

Contact Information | |
E-mail: | atersenov [at] physics [dot] uoc.gr |
Phone: | |
Office: | |
Affiliation: | IRFU/DAp-AIM |
Supervisors: | Jean-Luc Starck, Martin Kilbinger |
Research Interests
I am a third-year PhD student working at the intersection of cosmology, statistics, and machine learning, co-supervised by Jean-Luc Starck and Martin Kilbinger.
My research focuses on developing new statistical and data-driven methods to extract cosmological information from weak-lensing observations. In particular, I work on mass mapping, higher-order statistics (HOS), and simulation-based inference (SBI) — with the broader goal of making cosmological inference both more powerful and more trustworthy.
During my PhD, I have been:
-
Investigating how different mass-mapping algorithms affect cosmological parameter constraints.
-
Developing fast, uncertainty-aware weak-lensing reconstruction algorithms that combine physical forward models with deep-learning priors.
-
Using SBI to study the impact of baryonic feedback and other systematics on higher-order statistics and cosmological inference.
-
Working on validating theoretical predictions for wavelet-based weak-lensing statistics using large-deviation theory.
I am an active member of the UNIONS and Euclid collaborations, where I contribute to the mass-mapping component of the Euclid Level-3 (OU-LE3) pipeline and co-lead the Mass Mapping and Tomography brick of the HOWLS (Higher-Order Weak Lensing Statistics) project.
I am also co-leading a Euclid Key Paper on theory-based higher-order statistics to be applied to the first Euclid data release.
Beyond research, I am passionate about teaching and outreach. I have co-taught an Introduction to Data Science and Machine Learning course at the University of Crete and co-organized several international Astrostatistics Summer (2024 & 2025) and Winter Schools (2025), where I also lecture on statistical methods, machine learning, and data analysis for astronomy.
My current interests include:
-
Robust and interpretable simulation-based inference,
- Field-level inference and forward modeling for cosmological surveys
-
Multi-probe cosmology combining weak lensing, clustering, and CMB lensing,
-
Modern AI and generative models for cosmological data analysis,
-
Uncertainty quantification and systematics modeling in survey pipelines.
- Weak lensing mass-mapping
Education
Before starting my PhD, I completed my B.Sc. and M.Sc. in Physics at the University of Crete, where I became fascinated by the problem of connecting cosmological theory with real data. This naturally led me to my current work— with the overarching goal of making cosmological inference both more powerful and more trustworthy.
- Master of Science in Astrophysics & Space Physics, University of Crete, Greece, 2023
- Bachelor's degree in Physics, University of Crete, Greece, 2022
Publications
- Impact of mass mapping algorithms on cosmology inference
A. Tersenov, L. Baumont, M. Kilbinger, J.L. Starck Published in Astronomy & Astrophysics 698 (2025): A25. DOI | arXiv:2501.06961 - A plug-and-play approach with fast uncertainty quantification for weak lensing mass mapping
H. Leterme, A. Tersenov, J. Fadili & J.L. Starck Submitted to Astronomy & Astrophysics - Euclid preparation: Towards a DR1 application of higher-order weak lensing statistics
Euclid HOWLS Collaboration (S. Vinciguerra, ..., A. Tersenov, et al.) Submitted to Astronomy & Astrophysics arXiv:2510.04953