When HEP meets ML

High Energy Physics (HEP) has introduced Machine Learning (ML) techniques such as Multivariate Analysis and Neural Nets in the 90′s. Boosted Decision Trees are now the baseline for difficult tasks like the Higgs boson search.

However, the connections between the two research fields are poor. We are exploring ways to improve this in collaboration with the AppStat group of LAL and a few other ML experts.

The first milestone has been the organization of the very successful ATLAS Higgs Machine Learning Challenge in summer 2014 : all information can be found it the web site hosted by  LAL here and on  Kaggle.

Group member : David Rousseau.