EDUCADO-MWGaiaDN training school on astro-AI and ML

2 Mar 2026, 09:00 → 6 Mar 2026, 17:00 Europe/Brussels

Gent

Organized with financial support of

We are pleased to announce the EDUCADO–MWGaiaDN Training School on Astro–AI and Machine Learning, taking place 2–6 March 2026 in Ghent, Belgium.

 

Organized jointly by the EDUCADO and MWGaiaDN MSCA Doctoral Networks, this training school brings together leading experts in artificial intelligence, machine learning, high-performance computing, and data visualization. Over the course of five days, participants will take part in a structured programme of expert-led lectures and hands-on training sessions, providing ample opportunities for direct engagement with lecturers and peers, and fostering scientific exchange and collaboration across research environments and disciplines.

 

This training school is ideally suited for early-career researchers in astronomy and astrophysics, but is equally valuable for any scientist

interested in applying state-of-the-art AI/ML techniques to the exploration, analysis, and visualization of large observational,

experimental, or simulation-based datasets.

 

We look forward to welcoming you to Ghent for an enriching week of learning and collaboration.

The organisers

• Sven De Rijcke (UGent, Belgium) (LOC)
• Eric Muires (UGent, Belgium) (LOC)
• Nikos Gianniotis (HiTs, Germany)
• Johan Knapen (IAC, Spain)
• Anthony Brown (Univ. of Leiden, the Netherlands)
• Mario Jurić (DiRAC, USA)
• Merce Romero-Gomez (ICCUB, Spain)
• Shuyu Wang (Univ. of Leiden, the Netherlands)
• Konstantinos Zafeiropoulos (Univ. Coimbra, Portugal)
• Subhadeep Sarkar (ICCUB, Spain)
• Dahimar Sanchez (IAC, Spain)

Monday 2 March

Rubin Observatory and LSST: From Data to AI-assisted Discovery

This lecture and tutorial series offers a comprehensive introduction to Rubin Observatory and the Legacy Survey of Space and Time (LSST), equipping students with the skills needed to work with next-generation astronomical survey data. The lecture component covers key aspects of Rubin Observatory and the LSST, including the design of the observatory, its scientific objectives, the LSST Science Pipelines that process the images to produce science-ready data products, and using AI assistants to accelerate data exploration. The hands-on tutorial sessions complement the lectures by providing students with practical experience working with LSST and LSST-like data. Students will learn how to use the LSST Science Pipelines on real survey data, gaining experience in extracting scientific insights from large-scale datasets.

Conveners: Leanne Guy, Mario Juric

Coffee Break

Rubin Observatory and LSST: From Data to AI-assisted Discovery

This lecture and tutorial series offers a comprehensive introduction to Rubin Observatory and the Legacy Survey of Space and Time (LSST), equipping students with the skills needed to work with next-generation astronomical survey data. The lecture component covers key aspects of Rubin Observatory and the LSST, including the design of the observatory, its scientific objectives, the LSST Science Pipelines that process the images to produce science-ready data products, and using AI assistants to accelerate data exploration. The hands-on tutorial sessions complement the lectures by providing students with practical experience working with LSST and LSST-like data. Students will learn how to use the LSST Science Pipelines on real survey data, gaining experience in extracting scientific insights from large-scale datasets.

Conveners: Leanne Guy, Mario Juric

Lunch

Rubin Observatory and LSST: From Data to AI-assisted Discovery -- hands on

This lecture and tutorial series offers a comprehensive introduction to Rubin Observatory and the Legacy Survey of Space and Time (LSST), equipping students with the skills needed to work with next-generation astronomical survey data. The lecture component covers key aspects of Rubin Observatory and the LSST, including the design of the observatory, its scientific objectives, the LSST Science Pipelines that process the images to produce science-ready data products, and using AI assistants to accelerate data exploration. The hands-on tutorial sessions complement the lectures by providing students with practical experience working with LSST and LSST-like data. Students will learn how to use the LSST Science Pipelines on real survey data, gaining experience in extracting scientific insights from large-scale datasets.

Conveners: Leanne Guy, Mario Juric

Coffee Break

Rubin Observatory and LSST: From Data to AI-assisted Discovery -- hands on

This lecture and tutorial series offers a comprehensive introduction to Rubin Observatory and the Legacy Survey of Space and Time (LSST), equipping students with the skills needed to work with next-generation astronomical survey data. The lecture component covers key aspects of Rubin Observatory and the LSST, including the design of the observatory, its scientific objectives, the LSST Science Pipelines that process the images to produce science-ready data products, and using AI assistants to accelerate data exploration. The hands-on tutorial sessions complement the lectures by providing students with practical experience working with LSST and LSST-like data. Students will learn how to use the LSST Science Pipelines on real survey data, gaining experience in extracting scientific insights from large-scale datasets.

Conveners: Leanne Guy, Mario Juric

Welcome reception

Tuesday 3 March

Rubin Observatory and LSST: From Data to AI-assisted Discovery

This lecture and tutorial series offers a comprehensive introduction to Rubin Observatory and the Legacy Survey of Space and Time (LSST), equipping students with the skills needed to work with next-generation astronomical survey data. The lecture component covers key aspects of Rubin Observatory and the LSST, including the design of the observatory, its scientific objectives, the LSST Science Pipelines that process the images to produce science-ready data products, and using AI assistants to accelerate data exploration. The hands-on tutorial sessions complement the lectures by providing students with practical experience working with LSST and LSST-like data. Students will learn how to use the LSST Science Pipelines on real survey data, gaining experience in extracting scientific insights from large-scale datasets.

Conveners: Leanne Guy, Mario Juric

Coffee Break

Rubin Observatory and LSST: From Data to AI-assisted Discovery

This lecture and tutorial series offers a comprehensive introduction to Rubin Observatory and the Legacy Survey of Space and Time (LSST), equipping students with the skills needed to work with next-generation astronomical survey data. The lecture component covers key aspects of Rubin Observatory and the LSST, including the design of the observatory, its scientific objectives, the LSST Science Pipelines that process the images to produce science-ready data products, and using AI assistants to accelerate data exploration. The hands-on tutorial sessions complement the lectures by providing students with practical experience working with LSST and LSST-like data. Students will learn how to use the LSST Science Pipelines on real survey data, gaining experience in extracting scientific insights from large-scale datasets.

Conveners: Leanne Guy, Mario Juric

Lunch

Rubin Observatory and LSST: From Data to AI-assisted Discovery -- hands on

This lecture and tutorial series offers a comprehensive introduction to Rubin Observatory and the Legacy Survey of Space and Time (LSST), equipping students with the skills needed to work with next-generation astronomical survey data. The lecture component covers key aspects of Rubin Observatory and the LSST, including the design of the observatory, its scientific objectives, the LSST Science Pipelines that process the images to produce science-ready data products, and using AI assistants to accelerate data exploration. The hands-on tutorial sessions complement the lectures by providing students with practical experience working with LSST and LSST-like data. Students will learn how to use the LSST Science Pipelines on real survey data, gaining experience in extracting scientific insights from large-scale datasets.

Conveners: Leanne Guy, Mario Juric

Coffee Break

Rubin Observatory and LSST: From Data to AI-assisted Discovery -- hands on

This lecture and tutorial series offers a comprehensive introduction to Rubin Observatory and the Legacy Survey of Space and Time (LSST), equipping students with the skills needed to work with next-generation astronomical survey data. The lecture component covers key aspects of Rubin Observatory and the LSST, including the design of the observatory, its scientific objectives, the LSST Science Pipelines that process the images to produce science-ready data products, and using AI assistants to accelerate data exploration. The hands-on tutorial sessions complement the lectures by providing students with practical experience working with LSST and LSST-like data. Students will learn how to use the LSST Science Pipelines on real survey data, gaining experience in extracting scientific insights from large-scale datasets.

Conveners: Leanne Guy, Mario Juric

Wednesday 4 March

Machine Learning: From physics-Informed machine learning to machine-learning-informed physics, and back again.

Machine learning (ML) revolves, for a large part, about data. Yet in astronomy and astrophysics we often possess an additional, extremely valuable source of information: the governing physical equations. In this lecture, we describe how physics can inform machine learning, and conversely, how we machine learning can inform our physics simulations.
First, we introduce physics-informed approaches to model complex observations, such as ALMA spectral line images, and describe the successes and pitfalls of this approach.

Next, we reverse the perspective and show how machine learning methods, such as Gaussian Process inference, can be used, for instance, to improve solvers for differential equations.
This is known as probabilistic numerics, and we will demonstrate it with examples from radiation transport and numerical relativity.

Finally, we briefly discuss how this seemingly newfound marriage of theory and observations is not new, but provides an interesting formal framework to describe how we have been doing science in the past and will be doing it in the future.

Machine learning emulators are emerging as powerful tools for accelerating and enhancing astrophysical simulations. They are particularly powerful in the context of astrochemistry, where traditional chemical networks often involve hundreds of species and thousands of reactions and make simulations computationally expensive even under severely restrained spatial, temporal, and physical complexity. By providing fast and accurate approximations to these detailed chemical solvers, ML emulators offer a practical way to explore richer chemical–physical parameter spaces and to incorporate more realistic chemistry into simulations of star and planet formation, the ISM, and other astrophysical environments. In this lecture, I will introduce the principles behind ML-based emulators for chemistry, explaining how they are trained, validated, and deployed to replicate full chemical solvers at a fraction of the computational cost while retaining high accuracy. I will discuss current methodologies, their advantages and challenges, and show how they integrate into state-of-the-art astrophysical simulations.

Convener: Frederic De Ceuster

Coffee Break

Machine Learning: Improving theoretical models with ML

Machine learning (ML) revolves, for a large part, about data. Yet in astronomy and astrophysics we often possess an additional, extremely valuable source of information: the governing physical equations. In this lecture, we describe how physics can inform machine learning, and conversely, how we machine learning can inform our physics simulations.
First, we introduce physics-informed approaches to model complex observations, such as ALMA spectral line images, and describe the successes and pitfalls of this approach.

Next, we reverse the perspective and show how machine learning methods, such as Gaussian Process inference, can be used, for instance, to improve solvers for differential equations.
This is known as probabilistic numerics, and we will demonstrate it with examples from radiation transport and numerical relativity.

Finally, we briefly discuss how this seemingly newfound marriage of theory and observations is not new, but provides an interesting formal framework to describe how we have been doing science in the past and will be doing it in the future.

Machine learning emulators are emerging as powerful tools for accelerating and enhancing astrophysical simulations. They are particularly powerful in the context of astrochemistry, where traditional chemical networks often involve hundreds of species and thousands of reactions and make simulations computationally expensive even under severely restrained spatial, temporal, and physical complexity. By providing fast and accurate approximations to these detailed chemical solvers, ML emulators offer a practical way to explore richer chemical–physical parameter spaces and to incorporate more realistic chemistry into simulations of star and planet formation, the ISM, and other astrophysical environments. In this lecture, I will introduce the principles behind ML-based emulators for chemistry, explaining how they are trained, validated, and deployed to replicate full chemical solvers at a fraction of the computational cost while retaining high accuracy. I will discuss current methodologies, their advantages and challenges, and show how they integrate into state-of-the-art astrophysical simulations.

Convener: Camille Landri

Lunch

spheer.ai

Thursday 5 March

HPC & parallel coding: SLURM (Simple Linux Utility for Resource Management)

The course is designed to explore the utilization of modern
high-performance computing (HPC) infrastructures. Participants will
engage with essential tools and practices, including Git and VS Code for
version control, Slurm for job scheduling, and Python optimization
techniques. The curriculum consists of practical sessions to enhance
learning through hands-on applications, focusing on containerization and
workflow automation. By the end of the course, attendees will have a
foundational understanding of HPC environments and the skills to develop
and manage HPC applications effectively.

Convener: Bernd Doser

Coffee Break

HPC & parallel coding: Parallel coding

The course is designed to explore the utilization of modern
high-performance computing (HPC) infrastructures. Participants will
engage with essential tools and practices, including Git and VS Code for
version control, Slurm for job scheduling, and Python optimization
techniques. The curriculum consists of practical sessions to enhance
learning through hands-on applications, focusing on containerization and
workflow automation. By the end of the course, attendees will have a
foundational understanding of HPC environments and the skills to develop
and manage HPC applications effectively.

Convener: Bert Vandenbroucke

Lunch

HPC & parallel coding -- hands on

The course is designed to explore the utilization of modern
high-performance computing (HPC) infrastructures. Participants will
engage with essential tools and practices, including Git and VS Code for
version control, Slurm for job scheduling, and Python optimization
techniques. The curriculum consists of practical sessions to enhance
learning through hands-on applications, focusing on containerization and
workflow automation. By the end of the course, attendees will have a
foundational understanding of HPC environments and the skills to develop
and manage HPC applications effectively.

Convener: Bernd Doser

Coffee Break

HPC & parallel coding -- hands on

The course is designed to explore the utilization of modern
high-performance computing (HPC) infrastructures. Participants will
engage with essential tools and practices, including Git and VS Code for
version control, Slurm for job scheduling, and Python optimization
techniques. The curriculum consists of practical sessions to enhance
learning through hands-on applications, focusing on containerization and
workflow automation. By the end of the course, attendees will have a
foundational understanding of HPC environments and the skills to develop
and manage HPC applications effectively.

Convener: Bert Vandenbroucke

Friday 6 March

VO & visualisation: Virtual Observatory

VO: The Virtual Observatory (VO) is a set of APIs to find, access, and use astronomical data. If you have ever used TOPCAT or Aladin, you probably have used the VO, too. But since it is APIs, a natural place to deal with them is a library, and pyVO is one such library for python. In this tutorial, we will explore data discovery and TAP queries from pyVO, going from catalogues to images.

ARGOS: modern cosmological simulations rely on sophisticated assumptions and generate rich, highly complex datasets. Troubleshooting their operation and interpreting their results remains challenging. Existing visualization tools offer powerful capabilities but often come with a relatively high barrier for entry, and are not designed with a focus on knowledge discovery through interactive and intuitive visual analytics.

We present ARGOS, an open-source, web-based environment for real-time visual analytics, tailored to cosmological simulation outputs and designed with an emphasis on user experience. ARGOS combines GPU-accelerated browser rendering for interactive exploration of large datasets with a template-driven approach that enables rapid adaptation to various types of data and analysis workflows.

Convener: Markus Demleitner

Coffee Break

VO & visualisation: ARGOS: advanced visual analytics for simulations

VO: The Virtual Observatory (VO) is a set of APIs to find, access, and use astronomical data. If you have ever used TOPCAT or Aladin, you probably have used the VO, too. But since it is APIs, a natural place to deal with them is a library, and pyVO is one such library for python. In this tutorial, we will explore data discovery and TAP queries from pyVO, going from catalogues to images.

ARGOS: modern cosmological simulations rely on sophisticated assumptions and generate rich, highly complex datasets. Troubleshooting their operation and interpreting their results remains challenging. Existing visualization tools offer powerful capabilities but often come with a relatively high barrier for entry, and are not designed with a focus on knowledge discovery through interactive and intuitive visual analytics.

We present ARGOS, an open-source, web-based environment for real-time visual analytics, tailored to cosmological simulation outputs and designed with an emphasis on user experience. ARGOS combines GPU-accelerated browser rendering for interactive exploration of large datasets with a template-driven approach that enables rapid adaptation to various types of data and analysis workflows.

Convener: Paul Vauterin

Lunch

VO -- hands on

The Virtual Observatory (VO) is a set of APIs to find, access, and use astronomical data. If you have ever used TOPCAT or Aladin, you probably have used the VO, too. But since it is APIs, a natural place to deal with them is a library, and pyVO is one such library for python. In this tutorial, we will explore data discovery and TAP queries from pyVO, going from catalogues to images.

Convener: Markus Demleitner

Coffee Break