Stephen Roberts
Professorial Fellow; Professor of Machine Learning; Head of Machine Learning Research Group; Director of CDT in Autonomous Intelligent Machines and Systems
Stephen Roberts (FREng, FIET, MInstP, CEng, CPhys) is Professor of Machine Learning in the Department of Engineering Science.
He leads the Machine Learning Research Group and is Director of the Centre for Doctoral Training in Autonomous Intelligent Machines and Systems. His main research interests lie in the application and development of mathematical methods in data analysis and data-driven machine learning, in particular statistical learning and inference and their application to complex problems in science and engineering.
Recent research has focused on non-parametric Bayesian models for multi-sensor data fusion, system optimisation and network analysis. Particular emphasis is placed on the real-world applications of advanced theory and over many years he has applied these statistical methods to diverse problems in astrophysics, biology, finance and engineering as well embedding them in a variety of commercial and industrial settings.
Publications
Kieran Wood, Stephen Roberts, Stefan Zohren (2021).
Slow Momentum with Fast Reversion: A Trading Strategy Using Deep Learning and Changepoint Detection. https://arxiv.org/abs/2105.13727. Daniel Poh, Bryan Lim, Stefan Zohren and Stephen Roberts. (2021).
Enhancing Cross-Sectional Currency Strategies by Ranking Refinement with Transformer-based Architectures. https://arxiv.org/abs/2105.10019. Samuel Kessler, Vu Nguyen, Stefan Zohren, Stephen Roberts (2021).
Hierarchical Indian Buffet Neural Networks for Bayesian Continual Learning. Proceedings of UAI 2021. (to appear). https://arxiv.org/pdf/1912.02290.pdf. Aldo Pacchiano, Philip Ball, Jack Parker-Holder, Krzysztof Choromanski, Stephen Roberts (2021).
Towards Tractable Optimism in Model-Based Reinforcement Learning. Proceedings of UAI 2021. (to appear). https://arxiv.org/abs/2006.11911. Philip J. Ball, Cong Lu, Jack Parker-Holder, Stephen Roberts (2021).
Augmented World Models Facilitate Zero-Shot Dynamics Generalization From a Single Offline Environment. ICML (to appear). https://arxiv.org/abs/2104.05632. Kuok, S.C., and Roberts S.J. and Girolami, M. and Yuen, K.-V. (2021).
Broad Learning Robust Semi-active Structural Control: a Nonparametric Approach. Mechanical Systems and Signal Processing, (in press). https://authors.elsevier.com/c/1d5Qy39~t0Y0Ki. Wolfgang Fruehwirt, Leonhard Hochfilzer, Leonard Weydemann, Stephen Roberts (2021).
Cumulation, Crash, Coherency: A Cryptocurrency Bubble Wavelet Analysis. Finance Research Letters. Camilla Sterud, Signe Moe, Mads Valentin Bram, Stephen Roberts and Jan Calliess (2021).
Recurrent neural network structures for learning control valve behaviour. Automation, Robotics & Communications for Industry 4.0 (ARCI’ 2021) Alexander Camuto, Matthew Willetts, Brooks Paige, Chris Holmes and Stephen Roberts (2021).
Learning Bijective Feature Maps for Linear ICA. Proceedings of AISTATS 2021 (to appear). https://arxiv.org/pdf/2002.07766.pdf. Alexander Camuto, Matthew Willetts, Stephen Roberts, Chris Holmes, Tom Rainforth (2021).
Towards a Theoretical Understanding of the Robustness of Variational Autoencoders. Proceedings of AISTATS 2021 (to appear). https://arxiv.org/pdf/2007.07365.pdf. Matthew Willetts, Alexander Camuto, Tom Rainforth, Stephen Roberts, Chris Holmes (2021).
Improving VAEs’ Robustness to Adversarial Attack. Proceedings of ICLR 2021 (to appear), https://arxiv.org/abs/1906.00230 A. Aprem and S. Roberts (2021).
Optimal pricing in black box producer-consumer Stackelberg games using revealed preference feedback. Neurocomputing. (to appear)