James joined ICHEC in November 2020 as part of the Quantum Programming team and works towards developing applications and software on quantum computing platforms.
Prior to joining ICHEC, James received his PhD in Physics from Trinity College Dublin. The PhD centred around applying machine learning techniques to problems in condensed matter physics. Half of this research was based on using such techniques for the discovery of new materials, specifically: constructing a Curie Temperature predictor and using generative adversarial networks to produce novel molecules. The other half of the research consisted of using deep learning methods with many-body lattice models. Here, neural networks were found to be able to predict ground-state and finite-temperature properties exponentially faster than the exact solution.
“Machine learning density functional theory for the Hubbard model”, J. Nelson, R. Tiwari, S. Sanvito, Physical Review B 99 (7), 075132
“Predicting the Curie temperature of ferromagnets using machine learning”, J. Nelson, S. Sanvito, Physical Review Materials 3 (10), 104405