by Giles Strong

Below is a short summary of the IML workshop at CERN, which Markus Stoye has also reported on in the previous post.

Day 1 was a discussion with industry experts about the state and future of ML. In the afternoon there was work on the community white-paper that the IML plans to publish. This document is meant to be a road-map for where we want HEP to be in 10 years time with regards to ML. The proto-document is here: https://docs.google.com/document/d/1o9S0XE4ly4-LZh9y96DG-U7HH61m2cHboV6xSt1tI0g/edit#heading=h.l4vnrli7pz7r and we are encouraged to contribute to it. There’s also a dedicated Google group here: https://groups.google.com/forum/#!forum/hsf-cwp-machine-learning

Day 2 there were presentations from HEP experiments on flavour tagging algorithms using ML; some very nice results! The afternoon had short introductions on new TMVA developments and a Scikit Learn tutorial.

Day 3 had presentations of more general tagging algorithms as well as tutorials on Keras and R.

The agenda with the presentations may be found here: https://indico.cern.ch/event/595059/other-view?view=standard_inline_minutes

There was also a quark/gluon jet classification challenge being run throughout the workshop as part of an initiative to introduce ‘standard, public data-sets’ for comparing ML solutions (think HEP MNIST). Happy to report that yours truly came third, but further submissions are accepted. Send pull requests to https://gitlab.cern.ch/IML-WG/IMLWorkshop2017-Challenge, an explanation of the challenge and links to the data are also on that page.