Lloyd Fung

Welcome! I am a machine learning and fluid dynamics researcher with a strong interest in data assimilation in complex fluid problems, data-driven discovery of symbolic models and complex fluid phenomena. Currently, I am a Research Fellow at Imperial College London, working closely with Prof. Luca Magri at the Aeronautics department on scientific machine learning and its interface with traditional data assimilation and numerical methods. I am also a Fellow at the Schmidt Sciences funded AI in Science programme at the IX, an AI initiatives of Imperial.

my background

I completed an M.Eng. in Aeronautical Engineering at the Department of Aeronautics of Imperial College London in 2017, with a year in industry working at Rolls-Royce. Funded by the President’s scholarship of Imperial, I continued to pursue a PhD in the same department under the supervision of Dr Yongyun Hwang, and graduated in 2021 with a thesis on the continuum modelling of dilute active suspensions and their pattern formation. After that, I was elected as a Research Fellow at Peterhouse, during which I worked with Prof. Raymond Goldstein at the Department of Applied Mathematics and Theoretical Physics, Cambridge, on the fluid mechanics of cell motility in Stokes flow, collective dynamics of motile cells as a result of motility, and continuum modelling of dilute active suspensions and their pattern formation.

Lately, I have started gaining interest in applying data-driven techniques from recent advancements in machine learning to fluid and control problems. After a brief time working with the Alan Turing Institute and Prof. Matthew Juniper on developing numerical methods for assimilating data in large PDEs, I came back to Imperial Aeronautics as a Research Fellow (ICRF). My current research focuses on Bayesian model discovery and data assimilation of chaotic ODEs. I am also increasingly interested in applying information theory to study how the information structure embedded in physics/PDEs dictates numerical methods and inform the modelling and training in inverse problems. In doing so, I often find myself reflecting on the use of machine learning techniques in science and engineering.