This year the International Conference on Scientific Computing and Machine Learning (SCML) was held in Bath, from 14-17 September.

In recent years, machine learning methods for scientific computing have attracted much attention. Many methods are a combination of machine learning and/or theories of physics and/or computational mathematics.

The conference aimed to showcase the latest research in these areas, which have been fragmented while pursuing research in the same direction, to bridge the gap between them, and to promote collaboration.

Around 130 delegates joined the event coming from all over the world including the UK, Europe, China, Japan, India and the USA, and from industry as well as academic institutions.

Over the four days, delegates heard from 20 keynote and invited speakers, were able to choose from 52 contributor talks, attended two expert tutorials and viewed 22 posters. Lively discussions took place at the opening reception, during the conference dinner and in the breaks. We were also very happy to welcome exhibitor displays from Cambridge University Press, World Scientific Publishing and IOP Publishing.

In the feedback, delegates commented:

“Great colleagues, lots of young people, excellent selection of topics”

“Really nice community with lots of opportunity to make connections and get to know people. Perfect size for meeting people”

During the conference, the organising committee reviewed the posters and we offer our congratulations to the following for their winning submissions:

• Laurin Ernst and Sergey Dolgov (University of Bath), Univariate Compositional Normalizing Flows
• Ho Ching Lee and Julian Mak (Hong Kong University of Science and Technology), Goal-oriented adjoint-based optimisation of coastal protection layout
• Rikuto Fukushima (Stanford University), Masayuki Kano (Kyoto University), Kazuro Hirahara (RIKEN) and Makiko Ohtani (Kyoto University), Physics-informed deep leaning links geodetic data and fault friction 

The poster winners were able to choose their prizes from books generously donated by Cambridge University Press and World Scientific Press.

We now look forward to SCML 2027 which will take place in Trondheim, Norway.

This conference was supported by EPSRC Programme Grant on the Mathematics of Deep Learning (Maths4DL), JST CREST Prediction Mathematical Foundation, Operator Learning Based on Geometric Classical Field Theory and Infinite Dimensional Data Science, and by JST ASPIRE Deep scientific computing: integration of physical structure and deep learning through mathematical science.

The full schedule of talks can be viewed here: https://maths4dl.ac.uk/wp-content/uploads/2026/09/Conference_programme_SCML_2026_final.pdf