Machine Learning in Physics meeting by the CPG

The IoP had an inaugural Physics in the Spotlight event from 21st-25th October 2019, celebrating the move to their new head quarters in King’s Cross with events organised by many groups together. On the 24th of October the Computational Physics group (CPG) hosted a one-day meeting on machine learning applications in physics. This was in collaboration with the Particle Accelerator and Beam group, the Plasma Physics group and Polymer Physics group.

Attendees of the Machine Learning conference

The plenary sessions were a sell-out, hitting the 140 capacity of the lecture theatre in the new venue, showing the appetite and interest in the field. With keynote speakers from the Alan Turing Institute, STFC’s Scientific Machine Learning group, and an ex-accelerator scientist turned FinTech Machine Learning consultant who refreshed us on the journey machine learning had taken since his work with it on ion beam spectroscopy a decade ago, there was a strong and varied programme on forefront techniques. After a short break this was followed up by talks from Jacqui Cole (Cambridge) and Aldo Glielmo, the winner of the CPG thesis prize.

Opening keynote

The afternoon was split into two parallel sessions. This allowed us to explore developments more specific to each group interests. The CPG organized one with the plasma physics group and accelerator physics group getting together to understand commonalities in large facility design and data exploration. Presentations by Matt King and Hannah Kockelbergh and a panel discussion featuring Stephen Dann set the scene for a discussion session, which overall highlighted the need for cross community training to help those looking to exploit ML and data-centric methods for physics.

The second parallel session was organized with the Polymer Physics group and focused on machine including deep learning in soft and biological matter. Presentations related to Gaussian processes were given by Richard Graham (Nottingham) and Richard Clayton (Sheffield). Applications to 2D or 3D image data were by Alan Lowe (UCL), Rollo Moore and Ladislav Urban (NIHR); Dimitris Pinotsis (City/MIT) presentation was on network architectures.

Launch of the ‘Machine Learning: Science and Technology’ journal

The day also featured the launch of the IoP Journal ‘Machine Learning: Science and Technology’. Alongside the introduction to the scope of the journal, there was a cake cutting at the end of the day. Discussions continued over refreshments into the evening.

The organizing committee consisted of Jonathan Smith (CPG/PABG), Gavin Tabor (CPG), Bart Vorselaars (CPG), as well as Joao Cabral and Nigel Clarke (both Polymer PG) and Greg Daly (Plasma PG). Furthermore, David Dunning (PABG) was also helpful in finding some of the speakers.

2 Replies to “Machine Learning in Physics meeting by the CPG”

Leave a Reply to Andrei Zvelindovsky Cancel reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s