Physicist with specialist knowledge in machine learning, inverse problems, analytical error propagation, iterative algorithms, and neutron spectrum measurement; as well as a strong aptitude and a unique approach to problem-solving.

Research experience

PhD project: Modernising neutron spectrum unfolding for fusion applications—Sep 2023

• Developed a software to systematically select foils for activation foil neutron spectrum unfolding experiments, to replace the current ad hoc approach of foil selection. Oversaw the complete software development lifecycle, beginning from stakeholders engagement to software quality control.

• Derived more mathematically rigorous algorithms for unfolding, which were implemented into an unfolding code suite.

• Developed a new visualization technique for presenting gamma-ray spectra based on the statistics underlying their noise distributions (Poisson distribution).

• Performed an activation experiment, using GENIE to perform the nuclide identification and analysis, and then using the neutron spectrum unfolding suite developed at UKAEA to perform the unfolding.

• Collaborated with other organizations to perform neutron spectrum measurements and unfold their neutron spectra to satisfactory quality.

• Presented at various physics conferences, including the IoP and FuseNet.

Master’s project: Neutron spectrum unfolding suite developer (using Neural Networks)—Sep 2019

• Developed a novel approach to the long-standing problem of inferring the neutron spectrum in a underdetermined system.

• Collaborated with other programmers on the same repository using Git.

• Developed and optimized a Neural Network using Google’s TensorFlow framework, which is then incorporated into the unfolding suite along with the appropriate documentations.

Nuclear Physics Laboratory—Dec 2019, Oct—Dec 2018

• Identified unknown elements using neutron activation analysis and gamma spectroscopy using a High Purity Germanium detector.

• Measured the neutron flux in neutron bath using a BF₃ detector.

• Analysed the results using Python’s numerical processing capability; and then visualized them graphically via Python as well.

-0.5cm-1.2cm Systematic study of the effect of neighbours on the evolution of intragrain misorientation—Dec 2018

• Extracted the simulation results from Abaqus, a commercial thermomechanical modelling software, using Python; and subsequently optimized the extraction speed.

• Researched and investigated various mathematical methodologies for finding the average orientation, which is a previously un-explored problem in material science.

• Reported simulation result and comparison with results in literature in a paper.

• Gained theoretical understanding of and practical experience with Finite Element Modelling.

Nuclear Talent course on Machine Learning and Data Analysis for Nuclear Physics—Jul2020

• A 2 weeks course on machine learning techniques applicable to nuclear physics.

• Neural network: Universal approximation theorem, backprop, autoencoders, GANs

• Supervised learning techniques: Random forest, logistic regression

• Unsupervised learning techniques: PCA

• Reinforced learning

• Exercises include: Classifying the particle tracks found in a time projection chamber TPC, classifying the number and location of events in a detector

Fusor Group Project, University of Birmingham—Sep 2019

• Recruited as part of Fusor Project team, following from an interview selection process.

• Worked with 7 other more senior students to build a Farsworth nuclear fusor.

• Successfully applied for a grant of £2800 on behalf of the group to cover the cost of the power supply.

Software repertoire

• Programming languages: Python, Fortran, Bash, R, C++, PowerShell

• GIT: GitHub (@OceanNuclear) and GitLab(@OceanNuclear)

• Markdown language: LaTeX

• Propreitory software: Abaqus, Adobe Premiere Pro, Adobe InDesign, Vectr, Microsoft Excel, Word

Qualifications

PhD Physics, Sheffield Hallam University—Oct 2023

Title: ENFUSE: Effective Neutron Spectrometry for FUSion Environment

• Reviewed existing algorithms and developed new algorithms for neutron spectrum unfolding in underdetermined condition

• Selected foils to be used according their unfolding effectiveness and feasibility

• Expected to design a module for neutron activation foil irradiation inside fusion reactors as part of the degree.

-0.3cm-0.7cm MSc Physics and Technology of Nuclear Reactors, University of Birmingham—Oct 2019

Results: distinction; modules:

• Nuclear Instrumentation, Radiation Dosimetry and Protection

• Radiation Transport, Thermal Hydraulics and Reactor Engineering

• Reactor Materials, Reactor System and NDE

• Practical Skills

• Research Project

BSc Nuclear Science and Materials, University of Birmingham—Jul 2018

Results: 2:1; modules:

• Classical Mechanics and Relativity 1 & 2

• Electromagnetism I and Temperature and Matter (including Electric Circuits)

• Statistical Physics and Entropy

• Particles and Nuclei and Nuclear Physics

• Mathematics for Physicist 1A & 2

• Physics Laboratory 1 & 2

• Physics Communication Skills (including C++ Computing)

Shung Tak Catholic English College (Hong Kong)—Jul 2015

HKDSE (Hong Kong Diploma of Secondary Education exam)

• Physics 5**, English 5*, Algebra and Calculus 5, Chemistry 5, Mathematics 4, Biology 4

• Equivalent to A*A*AABB in A levels