Dr Michael Smith
MScs, PhD
School of Computer Science
Lecturer
Outreach, Open Days and Headstart Officer
Member of the Machine Learning research group
+44 114 222 1800
Full contact details
School of Computer Science
Regent Court (DCS)
211 Portobello
߲ݴý
S1 4DP
- Profile
-
Dr Michael Smith studied Computer Science at Warwick university, then, after a few years outside academia, joined Edinburgh to take MScs in Informatics and Neuroinformatics and a PhD in computational neuroscience, looking at where self-motion cues are processed and integrating, in the human brain (using fMRI).
After a bit of travelling he went to Kampala (Uganda) to lecture (in 2014) teaching AI to students at Makerere.
He is now a Research Fellow at the University of ߲ݴý in the department of Computer Science in the Machine Learning group. His work encompasses Differential Privacy and its applications to Gaussian process (GP) regression and classification, bounds on attacks to GP classifiers by adversarial examples, a kernel for regression over integrals and a method for tracking bees using retroreflective tags.
His work is in particular now focused on modelling air pollution in Kampala, using data from a network of low-cost sensors.
He is currently investigating probabilistically handling the calibration of the sensors using mobile units. This system will soon be incorporated into a pipeline providing live predictions for policy makers and stakeholders in the city.
- Research interests
-
- Gaussian Processes
- Air pollution
- Differential Privacy
- Machine Learning for International Development
- Bumblebee tracking
- Adversarial Examples/bounds using Gaussian Processes
- Publications
-
Journal articles
- . Biology Letters, 20(4).
- . Machine Learning.
- . Methods in Ecology and Evolution.
- Learning Nonparametric Volterra Kernels with Gaussian Processes. Advances in Neural Information Processing Systems, 29, 24099-24110.
- Differentially Private Regression and Classification with Sparse Gaussian Processes. JOURNAL OF MACHINE LEARNING RESEARCH, 22.
- . BMC Nephrology, 21(1).
- Machine Learning for a Low-cost Air Pollution Network.. CoRR, abs/1911.12868.
- Differentially Private Regression and Classification with Sparse Gaussian Processes.. CoRR, abs/1909.09147.
- Gaussian Process Regression for Binned Data.. CoRR, abs/1809.02010.
- Differentially Private Gaussian Processes.. CoRR, abs/1606.00720.
- Modelling calibration uncertainty in networks of environmental sensors. Journal of the Royal Statistical Society Series C: Applied Statistics.
Conference proceedings papers
- Nonparametric Gaussian Process Covariances via Multidimensional Convolutions. Proceedings of Machine Learning Research, Vol. 206 (pp 8279-8293)
- . Conference on Neural Information Processing Systems, 29 November 2022 - 1 December 2022.
- . 2020 25th International Conference on Pattern Recognition (ICPR), 10 January 2021 - 15 January 2021.
- Multi-task Learning for Aggregated Data using Gaussian Processes.. NeurIPS (pp 15050-15060)
- Multi-task Learning for Aggregated Data using Gaussian Processes. Proceedings of the Conference on Advances in Neural Information Processing Systems 32 (2019)
- Differentially Private Regression with Gaussian processes. Proceedings of the 21st International Conference on Artificial Intelligence and Statistics, Vol. 84 (pp 1195-1203). Playa Blanca, Lanzarote, Canary Islands, 9 April 2018 - 11 April 2018.
Working papers
- Differentially Private Gaussian Processes.
Preprints
- , arXiv.
- , arXiv.
- , arXiv.
- , arXiv.
- , arXiv.
- , arXiv.
- . Biology Letters, 20(4).
- Grants
-
Pollinator: Using Data Driven Artificial Intelligence to Reveal Pesticide Induced Changes in Pollinator Behaviour, BBSRC, 02/2024 - 08/2025, £321,811, as PI