The European Organization for Nuclear Research ( CERN ) has launched an online competition that invites interested parties to identify the Higgs boson among a large amount of data
The competition, which has been published on a well-known website , works as follows: participants are invited to process data from 800,000 particle collisions produced by a simulator of the Large Hadron Collider LHC.
The data can be divided into two categories: either those in which the Higgs particle appears to participate, or those where the Higgs does not participate. Each of the collisions is accompanied by 30 variables, such as the energy and direction of the particles that resulted from them.
Out of the 800.000 collisions, 250.000 have already been labeled as «Higgs» or «not Higgs» and based on this information, the contestants are asked to ‘train’ a computer program to distinguish the remaining ones between the two categories.
The participants then upload their results to the online site which evaluates them by compiling a ranking list. This machine learning competition is accompanied by cash prizes that in total exceed €10,000 and to date more than 1,100 teams have submitted entries.
The team that finishes first will also be invited to the CERN research center to discuss with the scientists about its solution and to be examined whether it could be adapted to the real research needs.
Machine learning plays an increasingly growing role in modern research and specific methods proved very useful for the discovery of the Higgs particle in 2012. The main characteristic is the «training» of a computer program based on a set of data, so that it can subsequently distinguish on its own what it must do.
«People love this kind of problems, they capture their imagination», said Jois Noa-Vanouke, analyst of the organizer of the website competition. “We nevertheless try to be the online home of data analysis”.
The competition so far has attracted a heterogeneous mix of people with different skills, and among the first participants so far are also individuals who have no professional relationship with physics. It will accept entries until the end of September.
