Event Details

• Date of Event: 31st March 2025 12:00 – 4th April 2025 14:00
• Location: Daresbury Laboratory, in person event
• Fee: £350 (covers 4 nights accommodation and catering) – if we secure additional sponsorship this fee will be reduced
• Pre-requisites: Students will be expected to bring their own laptop, to have a decent level of coding experience (see pre-requisites below) and provide a letter of support from their supervisor

Description

This machine learning for materials training course is being run by AIchemy Hub in collaboration with Physical Sciences Data Infrastructure (PSDI) initiative with support from STFC-SCD, PSDS, CCP5 and CCP9 as a follow up to the very popular 2023 Machine learning for Atomistic Modelling Autumn School. This training is targeted towards PhD students, in particular those in the Materials and Molecular Simulations field, who have experience of coding but are not highly experienced with machine learning. The aim of this training is to introduce attendees to the latest methods of machine learning for the atomistic simulation of materials.

This training will encompass a number of talks and practical sessions, focusing on the basics of machine learning, machine learning interatomic potentials and graph neural networks. There will also be the opportunity for attendees to present a poster on their work.

Learning outcomes

• Awareness of the state-of-the-art methods for machine learning for atomistic and molecular simulations
• Hands on experience of using machine learning for atomistic and molecular simulations

Outline Agenda – Draft

Sessions	Monday	Tuesday	Wednesday	Thursday	Friday
Morning		Descriptors	NNs	MLIPs general	Generative models
Afternoon	Intro to ML	Unsupervised ML	GNNs	MLIPs – molecules/ materials
Evening	Research talk – Kim Jelfs	Posters	BBQ	Research talk – Nong Arthrith

Pre-requisites

Students attending this course must already have a foundational level of Python experience and hands on experience of using Python in their research. You will be expected to provide your own laptop for the training course, although software installation will not be required. A letter of support will be required from your supervisor alongside your application, this will be requested by email following your application. This letter of support is to show the backing of your supervisor to attend the training and must be completed for your application to be assessed. 

Timelines & Fees

The application deadline is 1st December 2024. Supervisors will be contacted for a letter of support following you application. All letters of support must be submitted by 6th December.

You will be informed of the outcome of your application on 13th January 2025, you will have to accept your place within 1 week and payment is required by 16th February 2025. 

Food and 4 nights accommodation is included in the £350 fee paid for this event, travel to Daresbury (and public transport to /from the lab) is not included and will need to be covered by the attendee. If we are able to secure additional sponsorship for this event we will reduce the fee.

Please note: places on this course are limited and in the event of oversubscription to the training course we will favour a diverse group of attendees.  

Organising Committee

• Alin-Marin Elena, Scientific Computing Department STFC 
• Keith Butler, University College London
• Reinhard Maurer, University of Warwick 
• Kim Jelfs, Imperial College London 
• Alex Ganose, Imperial College London
• Ioan-Bogdan Magdău, Newcastle University
• Chris Mellor, Imperial College London
• Nicola Knight, University of Southampton