Machine learning algorithms have emerged as one of the driving forces for advancements in astroparticle physics and astronomy.
The future of this development can be expected to hold nothing less than groundbreaking scientific discoveries. The involved data analysis tasks, e.g. particle property reconstruction, particle classification and ML-based deconvolution, require a diverse set of algorithms, such as ensemble methods, regression, unsupervised learning, imbalanced learning and various types of neural networks. The efficient generation of large amounts of annotated examples in Monte Carlo simulations is a key characteristic of the field, as well as one of its largest challenges.
This interdisciplinary workshop will bring together leading experts from Computer Science, astroparticle physics and astronomy, to present and discuss recent developments at the intersection of these fast evolving fields. The ultimate goal of this workshop is to foster collaborations that span multiple disciplines to further advance all fields involved. We invite all submissions that are in line with this goal, including methodological, theoretical and practical works, as well as proofs of concepts. Submissions can address, without being limited to, the following topics:
- Unfolding / deconvolution / quantification
- Neural networks and graph neural networks (GNNs)
- Generative adversarial networks (GANs)
- Ensemble Methods
- Unsupervised learning
- Unsupervised domain adaptation
- Active class selection
- Imbalanced learning
- Learning with domain knowledge
- Particle reconstruction, tracking, and classification
- Monte Carlo simulations
Details regarding the dates, the submission, and the program are given below.
Program
This half-day workshop will host 1 physics keynote, 2 computer science keynotes, and oral presentations of all accepted papers.
Schedule
| Keynotes | |
| 14:30 – 15:00 | Katharina Morik, TU Dortmund University: Machine Learning for Astroparticle Physics |
| 15:00 – 15:30 | Daniel Nieto, Universidad Complutense de Madrid: Applications of Machine Learning to Gamma-Ray Astronomy |
| 15:30 – 16:00 | Alejandro Moreo, Instituto di Szienza e Tecnologie dell’Informazione: Quantification |
| 16:00 – 16:30 | break |
| Contributed Talks | |
| 16:30 – 16:50 | Mirko Bunse, TU Dortmund University: Unification of Algorithms for Quantification and Unfolding |
| 16:50 – 17:10 | Janis Kummer, Center for Data and Computing in Natural Sciences (CDCS)/ Universität Hamburg: Radio Galaxy Classification with wGAN-Supported Augmentation |
| 17:10 – 17:30 | Dmitry Malyshev, Erlangen Centre for Astroparticle Physics: To split or not to split classes of gamma-ray sources? |
| 17:30 – 17:50 | Shreyas Kalvankar, K. K. Wagh Institute of Engineering Education and Research: Astronomical Image colorization and up-scaling with Conditional Generative Adversarial Networks |
| 17:50 – 18:10 | Pranav Sampathkumar, Karlsruhe Institute for Technology: Sequential networks for cosmic ray simulations |
| 18:10 – 18:30 | Jigar Bhanderi, University of Erlangen-Nurenberg: Calculation of the Photon Flux in a Photo-Multiplier Tube with Deep Learning |
Keynote Speakers
Proceedings
All accepted papers will be published in the proceedings of INFORMATIK 2022. These proceedings appear in the series Lecture Notes in Informatics (LNI).
Submission
All submissions must adhere to the following requirements:
- The page limit is 6 to 10 pages, including references and appendices.
- All submissions must be formatted according to the LNI formatting guidelines.
- Each submission will be reviewed by at least two members of our program committee: one computer scientist and one physicist.
- Submissions are welcome in English and in German language.
Dates
All submission deadlines correspond to Anywhere-on-Earth (AoE) time.
| 26 Sep 2022 | Workshop (2pm – 6pm) |
Committee
The program committee of this workshop consists of renowned experts in astroparticle physics, machine learning, and astronomy. Clicking on one of the profiles will take you to their websites.
Contact: Any remaining question may be directed to Tim Ruhe or Dominik Elsässer.
