Bespoke Datasets

In addition to the general-use open data documented on this site, there are a number of bespoke datasets available in the CERN Open Data portal. These are mostly datasets generated with a specific research use-case in mind, like training a jet-tagging algorithm or a fast calorimeter simulation. For each of these datasets, documentation is provided in the CERN Open Data portal and the documentation and repositories that accompany the record.

ATLAS and ATLAS members have also contributed to setting up a number of "challenges," including the TrackML Particle Tracking Challenge, the Higgs Boson Machine Learning Challenge, and the FAIR Universe Higgs Uncertainty Challenge. While not all of these challenges use ATLAS (Open) Data directly, they attempt to build realistic datasets, often modeled on the ATLAS detector or Monte Carlo simulation, and can be useful for both research and educational purposes.

In case you believe some additional dataset would be of use to a wide community outside of the experiment, you are welcome to contact us. There are also opportunities to join the ATLAS for limited collaborations on specific projects.