Research
IceCube
Located at the South Pole, the IceCube Neutrino Observatory is the largest particle detector in the world, using thousands of light sensors buried in the Antarctic ice. It is at the forefront of neutrino and multi-messenger astronomy. One of the goals of IceCube is to identify astrophysical neutrino signals and their sources, and understand how they produce such high-energy particles.
A signal within IceCube is represented by a series of "bubbles" that show which sensors have detected light. These bubbles are color-coded, with red bubbles indicating the earliest light detected and bluer bubbles for the last detected light. The five main types of signals seen are: through-going track, starting track, stopping track, skimming track, or cascade.
Why IceCube Needs Your Help
IceCube consists of more than 5,000 light sensors arranged in a 3D grid one cubic kilometer deep in the South Pole ice sheet. One terabyte of raw data are recorded daily and we need your help to ensure we properly classify these events!
IceCube has recently developed machine learning algorithms that work to identify the different types of signals happening in IceCube. Unfortunately, the machine learning algorithms sometimes misclassify signals and we need your help to investigate this by looking at the signals individually. Your work will be compared to the performance of these machine learning algorithms.
After this comparison, we can begin work on two potential projects:
- Retrain the machine learning algorithm using your classifications.
- Switch to classifying real signals instead of simulations.
We are really hopeful that Name that Neutrino can help us uncover new ways of looking at the sky: instead of the light from the stars, we are seeing the particles.