Skip to content

Machine learning based score in InDet Ambisolver

Robin Newhouse requested to merge rnewhous/athena:21.0-trkml into 21.0

This implements a new machine-learning-based scoring function to order the priority of tracks to be reconstructed in dense tracking environments. This scoring function will be dormant for now and can be activated with an InDetFlag set in the jobOptions file. Tools developed to train and evaluate different network configurations will be stored in a separate repository in the Atlas-Inner-Tracking workspace.

Edited by Robin Newhouse

Merge request reports