BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//CERN//INDICO//EN
BEGIN:VEVENT
SUMMARY:Review of neural network methods for the BaikalGVD experiment
DTSTART;VALUE=DATE-TIME:20241023T150000Z
DTEND;VALUE=DATE-TIME:20241023T151200Z
DTSTAMP;VALUE=DATE-TIME:20260612T233303Z
UID:indico-contribution-4092@cern.ch
DESCRIPTION:Speakers: Grigory Plotnikov (MIPT)\nBaikal-GVD is a neutrino t
 elescope with an effective volume of approximately 1 km³\, located in Lak
 e Baikal. This experiment leverages a neural network-based approach to add
 ress multiple challenges in data analysis:\n\n 1. Suppression of noise hit
 s of the optical modules (OMs) caused by the natural luminescence of the m
 edium\, while preserving signal hits generated by Cherenkov radiation\;\n 
 2. selection of events caused by neutrinos against the dominant background
  of events caused by extensive air showers\;\n 3. Reconstruction of the ne
 utrino's arrival direction\;\n 4. Reconstruction of neutrino energy.\n\nMo
 nte Carlo simulations demonstrate that the developed neural networks achie
 ve performance metrics comparable to traditional methods. Futhermore\, for
  task 1\, the neural network surpasses standard techniques (achieving 99.5
 % "precision" metric versus 95%). For task 2\, a novel method developed th
 at estimates the total number of neutrino-induced events in a dataset and 
 the associated error. In task 4\, a neural network model is developed to p
 redict the energy of neutrino events along with an estimate of the predict
 ion error\, corresponding to one standard deviation.\n\nhttps://indico.par
 ticle.mephi.ru/event/436/contributions/4092/
LOCATION: Moskvorechye 2
URL:https://indico.particle.mephi.ru/event/436/contributions/4092/
END:VEVENT
END:VCALENDAR