For anyone looking for something fun to do over the summer months, while classes are out and many research projects are stalled for a variety of reasons, the people at Kaggle are hosting an interesting competition this summer.

The Higgs Boson Machine Learning Challenge is a contest to develop better methods of using machine learning methods, such as decision trees and neural networks, to improve the signal-to-noise ratio of particle physics experiments. Competitors will be provided with simulated data from the ATLAS experiment at the Large Hadron Collider, and must create a method of separating out reactions in which a Higgs boson decays into two tau leptons, from the huge number of background reactions that may appear to be similar but have slight differences. (Note though that it has been setup so that very little knowledge of physics is required, as the emphasis is on the machine learning aspects of the challenge)

Understandably, this competition has a very real application to future particle physics experiments, as reducing background signals is one of the biggest challenges in experimental physics. It runs from now until September 15, 2014, at which time the best algorithms/programs will receive up to $7000 in cash, and the methods that ATLAS deems most useful will also receive a partially paid trip to see the experiments themselves.

More details on the rules, requirements, and entry process can be found here.