Andrew Thompson leads the machine learning work of the Data Science department, where he focusses on uncertainty evaluation for machine learning. He also works on applications of compressed sensing in image reconstruction and quantum tomography.
Andrew joined NPL as a Senior Research Scientist in 2018 and he is also a Visiting Lecturer at the University of Oxford. He was previously Departmental Lecturer at the University of Oxford (2014-18) and Visiting Assistant Professor at Duke University (2012-14), advised by Robert Calderbank. He studied Mathematics at the University of Cambridge (2002) and obtained his PhD from the University of Edinburgh (2012), supervised by Coralia Cartis.
He has published research articles in high profile journals including Information and Inference, Scientific Reports, IEEE Transactions on Signal Processing, SIAM Journal on the Mathematics of Data Science and SIAM Journal on Optimization.
His current projects include
- Implementation of uncertainty evaluation for machine learning in various metrology applications.
- Current mapping of photovoltaic cells using compressed sensing reconstruction.
- Uncertainty-aware compressed sensing algorithms for quantum tomography.
- Sensor modelling and assurance of autonomous navigation systems.
- AI-enabled and uncertainty-aware future communications networks.