Navier-Stokes regularity, fluid computing, and machine learning workshop

This post is intended as an advertisement or extended abstract and open invitation for community of researchers who work at the interface of: (a) fluid mechanics; (b) machine learning and analog computing; (c) functional analysis of Navier-Stokes equations, to attend a series of talks in an upcoming workshop at the London Institute for Mathematical Sciences. The one-day workshop will be organised on June 1st from 11.30 AM to 5 PM.

The focus of the day will be on the Hardy lecture, to be delivered by Prof. Eva Miranda, who is visiting the UK from Barcelona to deliver a series of lectures, as part of the recent “Hardy lecture 2023” prize, conferred to her by London Mathematical Society. Her talk will be a semi outreach and semi technical talk on her result with Caronda, Peralta-Salas, and Presas that Euler equations on a Riemannian sphere $\mathbb{S}^3$ can be embedded in a deterministic Turing machine. After this opening talk, Dr Giulia Marcucci from Glasgow will present her recently constructed neuromorphic analog computers constructed from KdV and Nonlinear Schrodinger equations governing water waves. The third talk is by Dr Juan Gongora from Loughborough who will discuss nonlinear interactions among large number of waves in optical components that can be implemented in the form of logic gates or artificial neurons. A connected thread between Turing completeness of fluids and analog computing is the belief amongst computing community that the physical systems governed by equations which show singularity of certain kind have a capability of performing computations, supposedly because the non-unique weak solutions of these equations can encode the logic gates intended to perform analog computation (more papers in this directions are here).

After the remarkable connection between fluids and analog computing, the next focus of the workshop is on connection between fluids and machine learning. Michael will discuss recently developed machine learning methods used to accelerate the CFD algorithm processing times, especially for turbulent flow. This will be followed by Yang-Hui He’s lecture on application of ML algorithms in the field of algebraic geometry, representation theory, combinatorics, and number theory.

The workshop will finish with an art exhibition by Binghui and Tianxiang, recent graduates from the Royal College of Arts, who are going to present a Bismuth sculptures, which when projected on a screen, create visuals which are immersive in nature.

If you wish to register and attend this workshop, please fill in in your details here.
Travel support is available, and refreshments will be served.