iml-wg / HEP-ML-Resources

Listing of useful learning resources for machine learning applications in high energy physics (HEPML)
MIT License
331 stars 115 forks source link

Add Using holistic event information in the trigger arXiv:1808.00711 #30

Closed dtsbourg closed 6 years ago

dtsbourg commented 6 years ago

Hi 👋!

Making this PR to add the report on work I did as an intern at LHCb over the last 6 months. It covers the use of Machine Learning methods in Upgrade Trigger, and I'm sure it'll be of interest to the HEP-ML community :)

I has been released as a public LHCb note and is available on Inspire HEP or arXiv.

Cheers, Dylan

matthewfeickert commented 6 years ago

@dtsbourg Thanks for the PR --- looks nice! I see that there was a revision that was submitted on 2018-08-08 (hence you're using it for the data instead of the 2nd) but was it a big enough revision that we should depart from using the date submitted to INSPIRE? There is no answer I'm looking for here, just wondering what the difference is and if you have a preference (though I guess the 8th as that's what you submitted).

dtsbourg commented 6 years ago

The revision was due to a mistake on my part regarding the reference number (I had attributed it a mistaken CERN-EP number), but the content is equivalent. Either date is fine!

matthewfeickert commented 6 years ago

Cool. I'm going to then grab your branch and push an update to roll back the data to the 2nd. Congrats on the nice note and thanks for your contribution!