BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//CERN//INDICO//EN
BEGIN:VEVENT
SUMMARY:DORAEMON T1-WC
DTSTART:20260910T010000Z
DTEND:20260910T020000Z
DTSTAMP:20261005T133100Z
UID:indico-event-10751@indico.slac.stanford.edu
DESCRIPTION:https://u-tokyo-ac-jp.zoom.us/j/98126050083\nLink to recording
 \nQuick recap\nThe meeting focused on defining the data sets and challenge
 s for the T1 data challenge in water Cherenkov detectors. Kazuhiro emphasi
 zed the need to critically select data sets based on specific reconstructi
 on challenges\, rather than generating samples just because it's possible.
  He advised against mixing interesting studies with the core data challeng
 e and urged the team to start from scratch rather than using existing list
 s. Tashiro shared a tentative list of particle samples but was asked to re
 consider it based on the critical needs of Hyper-K. The discussion also co
 vered the validation of the LUCiD simulation's new electronics modeling an
 d the need for a baseline benchmark using fiTQun. Technical questions abou
 t metrics for multi-ring segmentation and the generation of samples for fi
 TQun tuning were addressed. The importance of involving experts like Ryo f
 or the baseline and ensuring the data sets are ready for an upcoming works
 hop was highlighted.\nNext steps\nOmar A. Alterkait\n\nImplement functiona
 lity in LUCiD to output per-photon information on scattering and reflectio
 n\, and to ensure readiness to generate "electron bomb" samples for fiTQun
  tuning.\n\nPatrick\n\nFollow up with the student (Mathias) working on mul
 ti-ring segmentation to explore metrics beyond the Dice score\, such as Pa
 noptic Quality (PQ)\, for handling ghost rings.\n\nRiya\n\nValidate the ne
 w electronics modeling in LUCiD simulation against SK models\, specificall
 y checking the PR made by Cesar.\n\nTakuya\n\nReconsider and propose a cur
 ated list of particle samples for the data challenge\, starting from scrat
 ch based on critical reconstruction challenges\, not just existing proposa
 ls.\n\nCollaboration\n\nPatrick and Takuya: Define the use case for each d
 ata set (training\, benchmarking\, physics application) and generate confi
 gurations accordingly.\nPatrick and Takuya: Communicate with Ryo to define
  the samples needed for fiTQun tuning and ensure their generation before t
 he workshop in Kyushu.\nPatrick and Takuya: Find a time slot to meet with 
 Cesar (US) to discuss data generation configurations.\nPatrick and Takuya:
  Consult with Hyper-K experts to prioritize the top reconstruction challen
 ges for the data challenge.\n\nSummary\nGenesis Sample Preparation Discuss
 ion\nTakuya discussed preparing a list of samples and a tentative option t
 able\, which is based on a proposal by Cesar but includes additional sampl
 es. Patrick inquired about clarifying the distinction between data used fo
 r training versus testing or benchmarking\, particularly for mu pi zero ap
 plications in physics\, such as proton decay.\nData Challenge Sample Gener
 ation Strategy\nThe team discussed sample generation for a data challenge\
 , with Kazuhiro emphasizing the need to have specific reasons for includin
 g each sample rather than generating data simply because it's possible. He
  argued that every sample should serve a unique purpose and be critically 
 evaluated\, distinguishing between smaller research studies and the main d
 ata challenge objectives. Kazuhiro stressed the importance of defining a c
 ommon benchmark that the community can use\, while avoiding the mixture of
  interesting but unrelated research studies with the main data challenge g
 oals.\nCritical Reconstruction Challenges Prioritization\nKazuhiro emphasi
 zed the need to focus on critical reconstruction challenges rather than bu
 ilding on existing work\, urging the team to generate new priorities from 
 scratch. Takuya agreed to reconsider the approach and start fresh for the 
 data challenge. The discussion highlighted the importance of identifying k
 ey challenges in areas like clustering\, multi-ring reconstruction\, PID\,
  and momentum reconstruction\, with experts being tasked to prioritize the
 se based on their expertise. Patrick mentioned Benda's interest in calibra
 tion and domain shift mitigation for reconstruction algorithms' robustness
  to detector configuration variations.\nFirst Release Focus and Validation
 \nKazuhiro emphasized that the current focus should be on completing the f
 irst release without expanding the scope\, and he urged the team to valida
 te the realistic electronic simulations claimed by Cesar for the dataset. 
 Patrick discussed the need to curate a list based on Hyper-K's needs\, par
 ticularly regarding machine learning and AI exploration. Takuya mentioned 
 considering data science activities and proposed making a proposal this we
 ek.\nElectronics Modeling Dataset Validation\nKazuhiro emphasized the need
  to validate the new electronics modeling and configure datasets based on 
 specific use cases\, such as benchmarking or training\, without using cost
  as a limitation. Patrick and Riya discussed ongoing work on validating LU
 CiD simulation and checking Cesar's implemented models. The team also cons
 idered generating common samples at the generator level for different dete
 ctors\, though Kazuhiro questioned its relevance for T1. It was decided th
 at the Japan team should prioritize data generation configurations and coo
 rdinate directly with Cesar for this task.\nNeutrino Reconstruction Approa
 ch Discussion\nKazuhiro explained that while Monte Carlo samples of neutri
 no interactions can be generated using particle bomb (a random combination
  of particles)\, the team typically avoids using neutrino event generators
  due to their known inaccuracies and biases. He clarified that the reconst
 ruction algorithm's primary focus should be on reconstructing visible part
 icles rather than attempting to reconstruct neutrino properties directly\,
  as this approach helps avoid learning biases from unreliable generators. 
 The discussion highlighted a disagreement between Kazuhiro and Takuya rega
 rding the approach to neutrino reconstruction\, with Kazuhiro advocating f
 or separating visible particle reconstruction from neutrino property infer
 ence.\nT2 Data Challenge and Multi-ring Metrics Discussion\nKazuhiro expla
 ined that the T2 data challenge is already addressing the task of modeling
  neutral properties using event generators with reconstructed particle kin
 ematics. Patrick raised a question about metrics for multi-ring segmentati
 on\, specifically regarding the use of Dice scores and the need for a bett
 er metric that can account for missed rings and fake "ghost rings" added b
 y algorithms. Kazuhiro confirmed that proposing new metrics like DICE is a
 cceptable but emphasized the need to discuss how metrics would work from b
 oth researcher and maintenance perspectives.\nDORAEMON Project Sample Opti
 mization\nThe team discussed generating particle samples and detector geom
 etries for the DORAEMON project\, with Kazuhiro emphasizing the need to fo
 cus on generic challenges rather than exact experiment replicas. They iden
 tified the need to optimize fiTQun as a baseline benchmark for water Chere
 nkov\, with Ryo agreeing to lead this effort and requiring electron bomb s
 imulations for fiTQun tuning. The team agreed to create a detailed list of
  required sample features\, including photon scattering and reflection inf
 ormation\, with Omar A. Alterkait confirming these modifications should be
  implementable in LUCiD. Kazuhiro stressed the importance of consulting wi
 th Ryo and preparing the lookup table before the upcoming meeting to ensur
 e productive workshop sessions.\n\nhttps://indico.slac.stanford.edu/event/
 10751/
URL:https://indico.slac.stanford.edu/event/10751/
END:VEVENT
END:VCALENDAR