Background
Before starting at compass, I graduated from Bristol university with a First Class honours Masters of engineering degree in Mathematics and Computer Science.
Prior to my PhD I did two big research projects.
- Bachelors dissertation: a project on simulating visual hallucination patterns in a spiking neural network in order to validate a field equation model.
- Masters dissertation: adapting and creating a novel extension to the Model-Agnostic-Meta-Learning (MAML) algorithm for temporal data to then perform k-shot learning on various tasks.
In the first year of my PhD alongside taught courses I undertook a mini-project where we examined the effect of pretrained learnt representations on the degree of catastrophic forgetting witnessed when performing a classification task in the continual learning setting.
Research Interests
With my supervisor Dr. Rihuan Ke I am currently working on self-supervised learning methods with a focus on contrastive methods and energy based models. We are carefully deriving new loss functions which can learn better general representations from data that can be useful for arbitrary tasks. Specifically, we are looking at loss functions that may be more helpful on currently more challenging problems, eg. multi-task, multi-modal, continual, online, etc. scenarios.
Check out my blogposts on the compass blog and check out my website for more info, links and additional blogposts: https://phd.h-0-0.com/