IPA workshop on Machine Learning for particle physics and astrophysics (IPA-ML)

Mar 20, 2023, 10:00 AM → Mar 22, 2023, 5:00 PM Europe/Zurich

HIT E 51 (ETH Zurich)

Annapaola de Cosa (ETH Zurich), Saul Alonso-Monsalve (ETH Zurich), Davide Sgalaberna (ETH Zurich)

The workshop, organised by the Institute for Particle Physics and Astrophysics (IPA) at ETH Zurich, aims to discuss diverse challenges in HEP and Astrophysics and how they are addressed using machine learning (ML). The event will consist of multiple talks from experts from academia and industry about successful applications of ML, together with some hands-on tutorials and discussions to provide the audience with sufficient knowledge on how to start profiting from these methods autonomously and making ML accessible to young researchers.

Monday, March 20 ¶

Methods: Introduction¶

1 Welcome¶

Speaker: Prof. Davide Sgalaberna (ETH Zurich)

Welcome IPA-ML.pdf

2 Physics and machine learning: an overview¶

Speaker: Dr Saul Alonso Monsalve (ETH Zurich)

ipaml-sam-23-03-20.pdf

12:00 PM Lunch

Methods: Machine learning and statistics¶

3 Machine learning lecture¶

Speaker: Dr Mauro Donega (ETH Zurich)

20230320_IPAML.pdf

2:45 PM Break

Tutorials¶

4 Machine learning (BDT) tutorial¶

Speakers: Simone Pigazzini (ETH Zurich), Massimiliano Galli

full notebook

intro_slides

Notebook on Colab

5 Deep learning tutorial¶

In this tutorial, you'll learn how to use deep learning to perform particle identification. We'll cover the basics of neural networks and how to build them using Python and TensorFlow. Participants will train their own neural networks using a dataset of particle tracks, and learn how to evaluate their performance. To participate, you'll need a Google account for Google Colab.

Prerequisites:

No prior experience with deep learning is required, but some familiarity with Python programming is recommended.

Materials needed:

A laptop computer with internet access and a Google account. Participants should open the notebook deep_learning_tutorial.ipynb on Google Colab from the following Github project: https://github.com/saulam/IPA-ML.

Speaker: Dr Saul Alonso Monsalve (ETH Zurich)

Tuesday, March 21 ¶

Applications: Neutrinos (I)¶

6 Introduction on neutrino experiments workflow with emphasis on challenges¶

Speaker: Prof. Davide Sgalaberna (ETH Zurich)

Neutrino Workflow.pdf

7 Image reconstruction in a 3D plastic scintillator detector using deep learning¶

Speaker: Dr Saul Alonso Monsalve (ETH Zurich)

ipaml-sam-23-03-21.pdf

8 Neutrino interaction classification and transfer learning¶

Speaker: Dr Leigh Whitehead (University of Cambridge)

lhw_zurich_machine_learning_200323.pdf

9 Event filtering and mitigation of simulation biases¶

Speaker: Dr Marta Babicz (University of Zurich)

Marta_slides.pdf

10:20 AM Break

Applications: Neutrinos (II)¶

10 Event reweighting and generative models in neutrino experiments¶

Speaker: Dr Cristovao Vilela (CERN)

cvilela_ipa-ml_20230321.pdf

11 Multi-task data reconstruction chain for imaging detectors in neutrino experiments¶

Speaker: Prof. Kazuhiro Terao (SLAC)

2023-03-21-ETH.pdf

2023_03_21_IPA_Kazu

Applications: Astrophysics (I)¶

12 Large Scale Structure Cosmology with Artificial Intelligence¶

Speaker: Dr Tomasz Kacprzak (ETH Zurich)

13 Cosmological constraints from combined probes of large scale structure with deep learning¶

Speaker: Arne Thomsen (ETH Zurich)

IPA_ml_arne_no_animations.pdf

12:10 PM Lunch

Applications: Astrophysics (II)¶

14 Background rejection in Cherenkov Telescopes¶

Speaker: Prof. Adrian Biland (ETH Zurich)

ML_IACT.pdf

15 New generative AI models¶

Speaker: Luca Biggio (ETH Zurich)

gen_models.pptx.pdf

16 Geometric Learning/ Physics-informed AI for cosmology¶

Speaker: Dr Tilman Troester (ETH Zurich)

geometric_deep_learning.key

geometric_deep_learning_tilman_troester.pdf

17 PointNets for galaxy redshift surveys¶

Speaker: Sotirios Anagnostidis (ETH Zurich)

PointNet for galaxy redshift surveys.pdf

2:40 PM Break

Applications: LHC experiments (I)¶

18 Introduction on LHC experiments workflow with emphasis on challenges¶

Speaker: Prof. Annapaola de Cosa (ETH Zurich)

IPAMLWorkshop_LHCPhyscis.pdf

19 Object identification and reconstruction¶

Speaker: Dr Alessandro Calandri (ETH Zurich)

ML_object_IPAworkshop_21032023.pdf

20 Ultrafast ML inference for triggering¶

Speaker: Dr Thea Aarrestad (ETH Zurich)

ipaml_triggering.pdf

21 Machine learning for data quality monitoring¶

Speaker: Roberto Seidita

ML4DQM.pdf

22 Probabilistic models in ML for HEP¶

Speaker: Davide Valsecchi (ETH Zurich)

23_03_21 - IPA ML workshop - Probabilistic models in ML for HEP.pdf

Wednesday, March 22 ¶

Applications: LHC experiments (II)¶

23 Data-MC matching¶

Speaker: Massimiliano Galli

230322_MLSeminars.pdf

24 Analysis techniques¶

Speaker: Florian Eble

analysis_techniques_HEP.pdf

25 Quantum machine learning¶

Speaker: Vasilis Belis (ETH Zurich)

qml_hep_belis.pdf

10:20 AM Break

Applications: Crossing boundaries in and beyond physics¶

26 Machine learning applications¶

Speaker: Dr Mauro Verzetti (Google)

[External] ML presentation @ ETH.pdf

27 Robotics, event-based imaging¶

Speaker: Prof. Davide Scaramuzza (ETH Zurich)

Scaramuzza_event_cameras.pdf

28 Knowledge distillation¶

Speaker: Patrick Odagiu (ETH Zurich)

Knowledge Distillation.pdf

12:30 PM Lunch

Methods: Round table and closing¶

29 Round table¶

30 Closing¶

Speaker: Prof. Annapaola de Cosa

IPAMLWorkshop_Closing.pdf