Deep-compression for High Energy Physics data - Honey Gupta - Google Summer of Code 2020 Project

Project has taken place from June 2020 to September 2020

Description: Storage is one of the main limiting factors to the recording of information from proton-proton collision events at the Large Hadron Collider at CERN. This project aims to implement an autoencoder based deep-compression algorithm on an ATLAS trigger machine to reduce this storage requirement and improve accessibility time of the collision data captured by a trigger system. We also plan to design an autoencoder model that has a better compression factor and adequate execution time and memory requirements in order to be deployed on a trigger system.

• Link to the Project

• Link to the Report

• Index of files uploaded

  • Initial proposal. Our project diverged a bit as we wanted to test event-level compression first.

  • Work log 

  • Meeting log

  • Timeline and deliverables

  • HTCondor instruction + GPU 

  • Full presentation 

  • Presentation for OpenLab

  • Processed data and description

  • All the plots with a README file containing the description of what the folders contain