BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//CERN//INDICO//EN
BEGIN:VEVENT
SUMMARY:RAMP 12: Solar storm detection
DTSTART:20181010T070000Z
DTEND:20181010T160000Z
DTSTAMP:20260907T154000Z
UID:indico-event-5048@indico.ijclab.in2p3.fr
DESCRIPTION:The goal of this RAMP is to detect Interplanetary Coronal Mass
  Ejections (ICMEs) in the data measured by in-situ spacecraft.\n\nICMEs ar
 e the interplanetary counterpart of Coronal Mass Ejections (CMEs)\, the ex
 pulsion of large quantities of plasma and magnetic field that result from 
 magnetic instabilities occurring in the Sun atmosphere (Kilpua et al. (201
 7) and references therein).  They travel at several hundred or thousands 
 of kilometers per second and\, if in their trajectory\, can reach Earth in
  2-4 days.\n\nICMEs interact with the planetary environment and may result
  in intense internal activity such as strong particle acceleration\, so-ca
 lled geomagnetic storms and geomagnetic induced currents. These effects ha
 ve serious consequences regarding space and ground technologies and unders
 tanding them is part of the so-called space weather discipline.\n\nICMEs s
 ignatures as measured by in-situ spacecraft thus come as patterns in time 
 series of the magnetic field\, the particle density\, bulk velocity\, temp
 erature etc. Although well visible by expert eyes\, these patterns have qu
 ite variable characteristics which make naive automatization of their det
 ection difficult. To overcome this problem\, Lepping et al. (2005) propose
 d an automatic detection method based on manually set thresholds on a set 
 of physical parameters. However\, the method allowed to detect only 60 % o
 f the ICMEs with a high percentage of false positives (60%). Moreover\, be
 cause of the subjectivity induced by the manually set threshold\, the meth
 od had difficulties to create a reproducible and constant ICME catalog.\n\
 nThis challenge proposes to design a machine learning algorithm to detect
  ICMEs from the most complete ICME catalog containing 657 events. We propo
 se to give to the users a subset of this large dataset in order to test an
 d calibrate their algorithm. We provide in-situ data measurement by the WI
 ND spacecraft between 1997 and 2016 that we sampled to a 10 minutes resol
 ution and for which we computed three additional features that proved to b
 e useful in the visual identification of ICMEs. Using an appropriate metri
 c\, we will compare the true solution to the estimation. The goal is to pr
 ovide an ICME catalog containing less than 10% of false positives while re
 cording as much existing event as possible. \n\nFormally\, each instance 
 consists of a measurement of various physical parameters in the interplane
 tary medium. The training set contains data measurement from 1997 to 2010 
 and the beginning and ending dates of the 438 ICMEs that were measured in 
 this period.\n\nYou will need to output the probability of being in an IC
 ME at each time step. We will compute various metrics on this prediction. 
 We will constrcut predicted ICMEs using this probability vector and score 
 you on the ICMEs of the 2010-2016 period using the recall obtained for a p
 recision of 90%. We will consider an ICME from our test catalog to be dete
 cted if it is overlapped by more than 50% by a predicted event.\n\nThe fol
 lowing image illustrates the challenge. \n\n\n\nVenue: INRIA Turing build
 ing\n\nhttps://indico.ijclab.in2p3.fr/event/5048/
LOCATION:Henri Poincaré (INRIA Turing building)
URL:https://indico.ijclab.in2p3.fr/event/5048/
END:VEVENT
END:VCALENDAR