Projects / Programmes
Preparing for dark matter search with the Cherenkov Telescope Array using machine learning
| Code |
Science |
Field |
Subfield |
| 1.02.03 |
Natural sciences and mathematics |
Physics |
Astronomy |
| Code |
Science |
Field |
| P002 |
Natural sciences and mathematics |
Physics |
| Code |
Science |
Field |
| 1.03 |
Natural Sciences |
Physical sciences |
dark matter, machine learning, gamma-ray astrophysics
Researchers (8)
Organisations (1)
| no. |
Code |
Research organisation |
City |
Registration number |
No. of publicationsNo. of publications |
| 1. |
1540 |
University of Nova Gorica |
Nova Gorica |
5920884000 |
14,070 |
Abstract
Over the past fifteen years, several satellites exploring the high-energy (HE) sky with gamma rays and charged cosmic rays gathered unprecedented data on the most energetic astrophysical processes in our Galaxy. These measurements resulted in a series of exciting discoveries which revolutionaries the field of HE astrophysics and provided an opportunity to search for a nature of Dark Matter (DM) particles in a completely new way. It was soon realized that one of the main limitations to taking the full advantage of this high-quality data are long computational times, needed for thorough explorations of parameter spaces. At the same time, machine-learning algorithms (MLAs) are being increasingly used to tackle the analyses of large data sets. They are now evolving to include elements of statistical analysis, which makes them suitable to address the scientific goals outlined above, and could bring to further breakthroughs in the field.
Improvements in the analysis of the HE data are also critical as the astrophysics community is preparing for the arrival of the next -generation ground based gamma ray observatory, the Cherenkov Telescope Array (CTA). It is the biggest gamma-ray experiment to date, included among the National infrastructural priorities and recently promoted to the Landmark status on the 2018 EU research infrastructure roadmap. The CTA is believed to be the only observatory capable of testing heavy thermal-dark matter models, which are among the most promising theoretical proposals to date. With its imminent arrival it is now timely to capitalize on what we learned from current HE observations (in particular from the gamma-ray satellite observatory, the Fermi LAT, which has comparable sensitivity but covers a lower energy range) and from the big-data analysis techniques to define a clear path forward for the new observatory.
The primary objective of this project is to develop novel data analysis techniques, by capitalizing on the knowledge gained from the Fermi LAT ten year mission and on the latest developments in the machine learning field, which will maximize the scientific output of the CTA in terms of dark mater search. The project leader (PL) is currently the coordinator of the working groups focused on dark matter of the Fermi LAT and CTA collaborations and a coordinator of a data challenge in the DarkMachines machine-learning platform. All project members have a track record of major contributions to this field, guaranteeing a successful project outcome.
Significance for science
The main goal of the proposed research is to develop advanced data analysis techniques which include the most promising approaches from the modern field of machine learning, to the field of gamma-ray astrophysics, focusing on the search for the nature of dark matter particles.
The project proposes an approach aimed at taking the full advantage of a previous experience of the project team in the ten years of data analysis and dark matter search with the satellite experiment Fermi LAT (which made a series of breakthrough discoveries in HE astrophysics and set some of the strongest limits on the nature of dark matter particles), in order to maximize the impact of a future experiment, the CTA, recognized not only as a top-priority project in the Astroparticle Physics European Consortium (APPEC) roadmap, but also national research infrastructure priority in Slovenia. By having the sensitivity comparable to that of the LAT, but covering the higher energies, the two instruments will share many of the data analysis challenges. In addition, the CTA will be the only experiment that can test heavy thermal dark matter, making our objective of the increased sensitivity even more pressing.
We stress that the machine learning techniques were never comprehensively implemented to the studies of the cosmic-ray interactions and dark matter search in our Galaxy (though smaller efforts showing the promise of this approach exist) and that the work here has a promise to open a brand new research direction in this field.
The project is therefore guaranteed to have a decisive impact in the field of HE astrophysics, still slow to introduce machine learning techniques while their benefits are demonstrated in other branches of modern life. In addition, our tests of applications of ground breaking techniques on the CTA data will be a critical step in the preparation for the upcoming observatory and will put our group and Slovenia on the fore-front of this growing field when the observatory comes online.
Significance for the country
The main goal of the proposed research is to develop advanced data analysis techniques which include the most promising approaches from the modern field of machine learning, to the field of gamma-ray astrophysics, focusing on the search for the nature of dark matter particles.
The project proposes an approach aimed at taking the full advantage of a previous experience of the project team in the ten years of data analysis and dark matter search with the satellite experiment Fermi LAT (which made a series of breakthrough discoveries in HE astrophysics and set some of the strongest limits on the nature of dark matter particles), in order to maximize the impact of a future experiment, the CTA, recognized not only as a top-priority project in the Astroparticle Physics European Consortium (APPEC) roadmap, but also national research infrastructure priority in Slovenia. By having the sensitivity comparable to that of the LAT, but covering the higher energies, the two instruments will share many of the data analysis challenges. In addition, the CTA will be the only experiment that can test heavy thermal dark matter, making our objective of the increased sensitivity even more pressing.
We stress that the machine learning techniques were never comprehensively implemented to the studies of the cosmic-ray interactions and dark matter search in our Galaxy (though smaller efforts showing the promise of this approach exist) and that the work here has a promise to open a brand new research direction in this field.
The project is therefore guaranteed to have a decisive impact in the field of HE astrophysics, still slow to introduce machine learning techniques while their benefits are demonstrated in other branches of modern life. In addition, our tests of applications of ground breaking techniques on the CTA data will be a critical step in the preparation for the upcoming observatory and will put our group and Slovenia on the fore-front of this growing field when the observatory comes online.
Most important scientific results
Interim report
Most important socioeconomically and culturally relevant results
Interim report
