2nd workshop on Artificial Intelligence for the Electron Ion Collider

US/Eastern
room 1019 (William & Mary, Raymond A. Mason School of Business, Alan B. Miller Hall )

room 1019

William & Mary, Raymond A. Mason School of Business, Alan B. Miller Hall

101 Ukrop Way, Williamsburg, VA 23185
Andrea Bressan (University of Trieste and INFN), Cristiano Fanelli (W&M), Markus Diefenthaler (Jefferson Lab), Tanja Horn (Cath), Torre Wenaus (BNL)
Description

This event follows the productive experience of the first AI4EIC workshop held in 2021 at CFNS and is organized by the EICUG AI WG. The scope of this second workshop is to cover all active and potential areas of applications of AI/ML for the EIC.

The workshop will include sessions on (i) accelerator and detector design (EPIC and potentially detector-2), (ii) connections to theory, (iii) analysis, (iv) reconstruction and particle identification, (v) infrastructure and frontiers in AI/ML and (vi) streaming readout, which will allow discussing different problems, perspectives and leading-edge solutions. 

During the workshop we will have AI/ML tutorial sessions provided by experts (academia, national labs, industry). The workshop will also host a Hackathon event (on October 14, whole day event), and a cash prize will be given to the solution winning the competition. 

During the first day, we will have talks with perspectives on AI/ML-related research from funding agencies. 

AI/ML will be an essential part of all phases of the future EIC and is already contributing to its realization starting from the design and R&D phases. This workshop is a great opportunity to update the community on the progress of ongoing projects and future plans, with discussions on multiple cross-cutting topics that bring together theorists, experimentalists, and AI/ML practitioners.

Live document is here

Instructions on meeting coordinates will be sent via email using the information provided in the registration form.

 

Location: 

William & Mary, Raymond A. Mason School of Business, Alan B. Miller Hall

Room 1019 

101 Ukrop Way, Williamsburg, VA 23185

Break-out rooms (available all time for AI4EIC participants) 1021-1022-1023

 

EICUG William & Mary Powered by AWS Cloud Computing

 

Participants
Contacts
    • 10:00 10:25
      Welcome & Intro to AI4EIC
    • 10:25 11:20
      Overview
      Conveners: Cristiano Fanelli (W&M), Tanja Horn (Cath)
      • 10:25
        EIC schedule and overview 40m
        Speakers: Elke-Caroline Aschenauer (BNL), Rolf Ent (Jefferson Lab)
      • 11:05
        Discussion 15m
    • 11:20 11:40
      Coffee break 20m
    • 11:40 12:30
      Opportunities for AI in NP and EIC
      Conveners: Cristiano Fanelli (W&M), Tanja Horn (Cath)
    • 12:30 13:30
      Lunch (on your own) 1h
    • 13:30 14:00
      Opportunities for AI in NP and EIC
      Conveners: Cristiano Fanelli (W&M), Tanja Horn (Cath)
      • 13:30
        NSF perspective on opportunities for AI in nuclear physics 30m
        Speaker: Bogdan Mihaila (National Science Foundation)
    • 14:00 17:00
      AI/ML for Design
      Conveners: Evaristo Cisbani (Italian National Institute of Heath and Italian Institute of Nuclear Physics), Wouter Deconinck (University of Manitoba)
      • 14:00
        Tutorial on Ax/Botorch: Multi-Objective Optimization 45m
        Speaker: Max Balandat (Meta/AI Facebook )
      • 14:45
        Discussion 10m
      • 14:55
        Adaptive experimentation to assist detector design at EIC: from ECCE to ePIC and beyond 20m
        Speaker: karthik suresh (University of Regina)
      • 15:15
        Q/A 5m
      • 15:20
        AI-driven detector design for the EIC 20m
        Speaker: Benjamin Nachman
      • 15:40
        Q/A 5m
      • 15:45
        Coffee Break 15m
      • 16:00
        ML application for beam optics control in the LHC 20m

        Particle accelerator optimization problems deal with non-linear,
        multi-objective functions which depend on thousands of time-varying machine components and settings. These properties often meet the limitations of traditional optimization methods and make this problem a perfect candidate for
        application of ML-based techniques. In this talk I will present, how ML can improve the control of the beams on the example of the LHC and give a short outlook on the ML application to accelerator design. Main focus of the presentation will be the application of decision tree - based methods to instrumentation faults detection, reconstruction and correction of magnet
        errors, and supervised learning for virtual diagnostics, which enables to obtain accurate information of beam properties without time-costly measurements.

        Speaker: Elena Fol (CERN)
      • 16:20
        Q/A 5m
      • 16:25
        AI/ML overview for accelerator design activities 20m
        Speaker: Todd Satogata (Jefferson Lab)
      • 16:45
        Q/A 5m
      • 16:50
        Discussion 10m
    • 10:00 13:00
      Experiment/Theory Connections
      Conveners: Maria Zurita (University of Regensburg), Markus Diefenthaler (Jefferson Lab)
    • 13:00 14:00
      Lunch (on your own) 1h
    • 14:00 17:00
      Experiment/Theory Connections
      Conveners: Benjamin Nachman, Justin Stevens (William & Mary)
    • 10:00 13:00
      Reconstruction and Particle Identification
      Conveners: Anselm Vossen (member@duke.edu;faculty@duke.edu), Rachel Montgomery, Sadhana Dash
      • 10:00
        Interpretable Networks for Identifying Leptons 25m
        Speaker: Daniel Whiteson (UC Irvine)
      • 10:25
        Q/A 5m
      • 10:30
        Tagging heavy flavor jets @ RHIC 25m
        Speaker: Raghav Kunnawalkam Elayavalli (Vanderbilt University)
      • 10:55
        Q/A 5m
      • 11:00
        Muon Identification with Deep Learning at EIC 15m
        Speaker: William Phelps (Christopher Newport University/Jefferson Lab)
      • 11:15
        Q/A 3m
      • 11:18
        Coffee break 10m
      • 11:28
        Machine Learning in ACTS 17m
        Speaker: Corentin Allaire (IJCLAB)
      • 11:45
        Q/A 3m
      • 11:48
        ML particle identification with measured shower profiles from calorimetry 15m
        Speaker: Chao Peng (Argonne National Laboratory)
      • 12:03
        Q/A 3m
      • 12:06
        Lambda event tagging at CLAS12 15m
        Speaker: Matthew McEneaney (Duke)
      • 12:21
        Q/A 3m
      • 12:24
        ML for calorimetry 15m
        Speaker: Nathan Branson (Messiah U.)
      • 12:39
        Q/A 3m
      • 12:42
        Data-driven learning: Flux+Mutability 15m
        Speaker: James Giroux (U. Regina)
      • 12:57
        Q/A 3m
    • 13:00 14:00
      Lunch (on your own) 1h
    • 14:00 17:00
      Infrastructure and Frontiers
      Conveners: Gabriel Perdue (Fermilab), Joe Osborn (Brookhaven National Laboratory), Malachi Schram (Thomas Jefferson National Accelerator Facility), Yihui Ren (BNL)
      • 14:00
        Tutorial on MLFlow 45m
        Speaker: Kishansingh Rajput (Jefferson Lab)
      • 14:45
        Q/A 5m
      • 14:50
        Coffee break 10m
      • 15:00
        Foundation Model Infrastructure 13m
        Speaker: Svitlana Volkova
      • 15:13
        Q/A 5m
      • 15:18
        AI/ML hardware co-design 13m
        Speaker: Frank Liu (ORNL)
      • 15:31
        Q/A 4m
      • 15:35
        Machine Learning with FPGA 13m
        Speaker: Nhan Tran (Fermilab)
      • 15:48
        Q/A 4m
      • 15:52
        AI/Ml with HPC 13m
        Speaker: Joo Balint (ORNL)
      • 16:05
        Q/A 5m
      • 16:10
        break 5m
      • 16:15
        Panel Discussion 45m
        Speakers: Jin Huang (Brookhaven National Lab), Mike Williams (MIT), Tia Miceli (Fermilab)
    • 09:00 17:00
      Hackathon
      Conveners: Cristiano Fanelli (W&M), Diana McSpadden (Jefferson Lab), Kishansingh Rajput (Jefferson Lab), Wouter Deconinck (University of Manitoba)