The GRADnet Machine Learning and AI Workshop took place on Wednesday, 20 November for SEPnet postgraduate and postdoctoral researchers at ISH Venues in London. 

Delegates were given short lectures on the landscape of machine learning and the different types of algorithms and looked at the interplay between industry and experimental particle physics development of algorithms and their use of them. Through the day students learned details about supervised learning, random forests, decision trees and Neutral Networks.

The online tutorials using Google Collab introduced random forest, decision tress and NN. Students developed a simple NN algorithm using Scikit;  learned to analyse proton proton collision data from the ATLAS experiment at the LHC to try to find a very rare event (Dark Matter).

The discussion sessions looked at various and more broad areas of ML, discussing how they are used in physics and in industry, the differences, the drawbacks and ethical issues, what types of algorithms are best for what types of problems.

Thank you to our presenters Dr Thomas Stevenson; Ms Caley Yardley and Ms Zoe Earnshaw, University of Sussex/DISCnet.

Attendees found the training really useful and enjoyed the hands-on element.