The HL-LHC will see ATLAS and CMS see proton bunch collisions reaching track multiplicity up to 10.000 charged tracks per event. Algorithms need to be developed to harness the increased combinatorial complexity. To engage the Computer Science community to contribute new ideas, we have organized a Tracking Machine Learning challenge (TrackML). Participants were provided events with 100k 3D points, and are asked to group the points into tracks; they are also given a 100GB training dataset including the ground truth.
The TrackML challenge has run in two phases.
- The first "Accuracy" phase has run on Kaggle platform from May to August 2018; algorithms were judged judged only on a score related to the fraction of correctly assigned hits. The first phase has seen 653 participants, with top performers with innovative approaches, exposed at NeurIPS 2018 Competition workshop and documented in 1904.06778
- The second "Throughput" phase ran Sep 2018 to March 2019 on Codalab, required code submission; algorithms were then ranked by combining accuracy and speed. It has seen astonishingly fast submissions.
This workshop will gather at CERN organizers, top competitors and anyone with interest with fast tracking.
Registration is free.
Non CERN user are welcome (they need to tick the appropriate box in the registration form).
The workshop takes place from Monday 1:30PM to Tuesday 5:30PM. Tuesday morning is reserved for visit of CERN highlights for non HEP participants.In particular, visits of ATLAS or CMS experiments which will be exceptionally open are foreseen. (Visits are fully booked, new registrants will be on the waiting list).
Contact : trackml-contact@googlegroups.com
