Working Thesis Title
Optimization Methods with Applications to Large Scale Machine Learning Problems.
The purpose of my PhD thesis is to study algorithms for large scale optimisation problems arising from applications to machine learning. Ideally, we would like to develop algorithms that are scalable to very high dimensional settings, and that are also able to locate (the) solution(s) efficiently. We can split these algorithms into two categories. There are first-order methods that are very cheap but require lots of iterations to converge, and second-order methods that are expensive but take many fewer iterations to converge. We will search for a middle ground, that is, first-order algorithms that have partial second-order information to help locate a solution quickly.
Supervisors:
Primary Supervisor: Rachael Tappenden
Research Interests
Optimization, Linear Algebra and Numerical Methods.
Academic History
I completed a BSc in Mathematics and Computer Science at UC in 2020, then completed my Honours degree in Mathematics in 2021.
