Oliver Sutton

Biography

Oliver Sutton is an applied mathematician specialising in Machine Learning and Numerical analysis. Key focuses of his current work include the challenging problem of understanding algorithms which learn from very limited data, and exploring the fundamental stability and instability properties of neural network classifiers learned from data.

Publications

Here are some recent relevant publications:

• Q. Zhou, O. J. Sutton, Y. Zhang, A. N. Gorban, V. A. Makarov, and I. Y. Tyukin (2023). Neuromorphic tuning of feature spaces to overcome the challenge of low-sample high-dimensional data, International Joint Conference On Neural Networks (IJCNN) 2023.
• I. Y. Tyukin, O. J. Sutton, and A. N. Gorban (2023). Learning from few examples with nonlinear feature maps, Science and Information Conference (2023) vol 711, pp 210-225, Springer, Cham. Winner of best paper award.
• O. J. Sutton, A. N. Gorban, and I. Y. Tyukin (2023). A geometric view on the role of nonlinear feature maps in few-shot learning, Geometric Science of Information (GSI) 2023 vol 14071, pp 560-568, Springer, Cham.
• O. J. Sutton, Q. Zhou, A. N. Gorban, and I. Y. Tyukin (2023). Relative intrinsic dimensionality is intrinsic to learning, International Conference on Artificial Neural Networks (ICANN) 2023 vol 14254, pp 516-529, Springer, Cham.
• A. Bastounis, A. N. Gorban, A. C. Hansen, D. J. Higham, D. Prokhorov, O. J. Sutton, I. Y. Tyukin, and Q. Zhou (2023). The boundaries of verifiable accuracy, robustness, and generalisation in deep learning, International Conference on Artificial Neural Networks (ICANN) 2023 vol 14254, pp 530-541, Springer, Cham.
• O. J. Sutton, Q. Zhou, I. Y Tyukin, A. N. Gorban, A. Bastounis, and D. J. Higham (2023). How adversarial attacks can disrupt seemingly stable accurate classifiers, arXiv 2309.03665.
• O. J. Sutton, A. N. Gorban, and I. Y. Tyukin (2022). Towards a mathematical understanding of learning from few examples with nonlinear feature maps, arXiv 2211.03607.