AI and ML Applications in Neutrino Astronomy and Cosmic Ray Composition Measurement
Explore the current use and future potential of artificial intelligence and machine learning in neutrino astronomy, focusing on the KM3NeT group's research objectives, involvement in AI/ML advancements, and plans for improvements in reconstruction and classification. Discover how AI may revolutionize acoustic neutrino detection and address the challenges in cosmic ray composition measurement within the next 5-10 years, highlighting the need for education and expertise development at Nikhef.
Download Presentation
Please find below an Image/Link to download the presentation.
The content on the website is provided AS IS for your information and personal use only. It may not be sold, licensed, or shared on other websites without obtaining consent from the author. If you encounter any issues during the download, it is possible that the publisher has removed the file from their server.
You are allowed to download the files provided on this website for personal or commercial use, subject to the condition that they are used lawfully. All files are the property of their respective owners.
The content on the website is provided AS IS for your information and personal use only. It may not be sold, licensed, or shared on other websites without obtaining consent from the author.
E N D
Presentation Transcript
http://www.km3net.org/logo/oldLogo/KM3NeT_logo_web_no_shade.gifhttp://www.km3net.org/logo/oldLogo/KM3NeT_logo_web_no_shade.gif AI/ML in the KM3NeT* group Aart Heijboer * Asked also Dune, but nikhef-group is hardware-focussed, for now 1
http://www.km3net.org/logo/oldLogo/KM3NeT_logo_web_no_shade.gifhttp://www.km3net.org/logo/oldLogo/KM3NeT_logo_web_no_shade.gif Current situation production use of AI/ML Production use of AI/ML: Currently no (advanced) AI/ML in production (!) Reconstruction = classical Maximum Likelihood, classification: BDT* Outside production (mostly outside Nikhef) GNN, CNN for reconstruction and classification; performance promising, but worries about changing/moving detector In-house experts : Not many (Ronald Bruijn probably most expert ) * BDT and other feature- based MVA s are used frequently, but don t count as AI/ML Several Msc projects MSc (Leiden, 2024) Jesse Bosman (background/track/shower discrimination) MSc (Leiden, 2021) Gijs Vermarien (tau identification / track-shower via GNN) MSc (Delft, 2023), Liselotte Dijkema (accoustic neutrino detection). 2
http://www.km3net.org/logo/oldLogo/KM3NeT_logo_web_no_shade.gifhttp://www.km3net.org/logo/oldLogo/KM3NeT_logo_web_no_shade.gif Mid-term Expected improvements on 3 yr timescale Improve reconstruction and classification of neutrinos near the detection threshold (lower threshold for astronomy, and improve mass hierarchy measurement) In particular: Address wide variety of (hadronic) final states in ORCA Use ML/AI to measure the composition of Cosmic Rays mixed AI/Maximum likelihood reconstruction ( let AI model the PDF ) Nikhef involvement: currently not leading institute in AI/ML (in fact, we lead in the more classical approaches). Leading institute: Ecap, Erlangen. Ambition/needed resources/etc @Nikhef group in next years to be clarified. Sentiment is we should get into the business, though. First step seems (self)education. Expectorations/plans for coding assistance: No formal plan, but some of use this and we expect it will be important in future. 3
http://www.km3net.org/logo/oldLogo/KM3NeT_logo_web_no_shade.gifhttp://www.km3net.org/logo/oldLogo/KM3NeT_logo_web_no_shade.gif Long-term future (5-10 years) future directions Acoustic neutrino detection: Nikhef group doing R&D in future acoustic neutrino detector AI may be enabling technology to find acoustic signals in huge background. Is there any other information regarding the use and potential of ML/AI in your research area that is relevant to be discussed in the task force? Basic knowledge and expertise is missing; propose education of Nikhef staff. 4
