Bio
Matthew has been a lecturer in Applied Mathematics at the University of Manchester since March 2020. Previously he was a Research Fellow at the University of Cambridge (September 2017- February 2020) and a Postdoctoral Associate at Carnegie Mellon University (September 2015 - August 2017). He obtained his PhD from the University of Warwick in 2015.
Research interests
Matthew studies theoretical properties of machine learning algorithms from the perspective of applied analysis. In particular, he works on large data limits of semi-supervised algorithms on graphs and applications of optimal transport distances. Typical research questions include: what fraction of data should be labelled for an asymptotically (as the total number of data points) goes to infinity and how robust is the methodology. These questions rely on the regularity of solutions to variational problems. The methods involved in answering these questions draw on theory from the calculus of variations, PDEs, probability and statistics. Matthew has worked on applications from imaging, medicine and materials working with clinicians, engineers and physicists.