Inference group

Advanced novel statistical methodology with diverse applications in science and engineering.

The Inference Group

The Inference Group at Imperial is part of the Statistics Section of the Mathematics Department and is led by Professor Mark Girolami.  The activities of the group covers the investigation and development of advanced novel statistical methodology, driven by applications in the physical, chemical, engineering and socio-economic sciences.

For a full list of research activities please visit the projects page.  Please contact Professor Mark Girolami if you have any queries or are interested in collaborating.

News

  • June 2018: Alex Terenin was awarded the prize for the best poster at the Imperial College London SIAM Student Conference.
  • June 2018: Our group has been acknowledged by the STARTS Prize winners for collaborations on the world’s first 3-D printed metal bridge project.
  • April 2018: Our paper Stochastic modelling of urban structure has been published in Proceedings of the Royal Society A.
  • April 2018: Mark Girolami to lead a five-year project Bridging big data and engineering.
  • April 2018: Mark Girolami has been appointed the Lloyd’s Register Foundation / Royal Academy of Engineering Research Chair in Data-Centric Engineering.

Links

Video Presentations

Data-centric engineering. Hype or hope?

Data-centric engineering. Hype or hope?Play video
Data-centric engineering. Hype or hope?

Probabilistic Numerics: A New Concept?

Probabilistic Numerics: A New Concept?Play video
Probabilistic Numerics: A New Concept?

https://www.youtube.com/watch?v=VrIMLt__CJc

The Future of Crime DetectionPlay video
Future of Crime Detection and Prevention
 
 
 
 
 
 
 
 
 
  

Recent Publications

L. Ellam, G. Pavliotis, M. Girolami, A. Wilson (2018), Stochastic modelling of urban structure, Proc. R. Soc. A, rspa.2017.0700, 2018.

Xi, X., Briol, F-X. & Girolami, M. (2018). Bayesian Quadrature for Multiple Related Integrals. International Conference on Machine Learning, arXiv:1801.04153.
 
Chen, W. Y., Mackey, L., Gorham, J. Briol, F-X. & Oates, C. J. (2018). Stein Points. International Conference on Machine Learning (to appear), arXiv:1803.10161.
 
Chris J. Oates; Jon Cockayne; F-X Briol; Mark Girolami (2018), Convergence Rates for a Class of Estimators Based on Stein’s Identity, Bernoulli, arXiv:1603.03220.
 
Barp, A., Briol, F. X., Kennedy, A. D., & Girolami, M. (2017). Geometry and Dynamics for Markov Chain Monte Carlo. Annual Review of Statistics and Its Application,
 
C. J. Oates, S. Niederer, A. Lee, F-X. Briol & M. Girolami (2017). Probabilistic Models for Integration Error in the Assessment of Functional Cardiac Models. arXiv:1606.06841. Accepted for publication at "Advances in Neural Information Processing Systems" (NIPS).
 

L. Ellam, H. Strathmann, M. Girolami, I. Murray (2017). A determinant-free method to simulate the parameters of large Gaussian fields. Stat DOI: 10.1002/sta4.153.

F-X Briol, C.J. Oates, J. Cockayne, W.Y. Chen, and M.A. Girolami (2017). On the Sampling problem for Kernel Quadrature. International Conference on Machine Learning (ICML).

K. Jensen, C. Soguero-Ruiz, K.O Mikalsen, R-O. Lindsetmo, I. Kouskoumvekaki, M. Girolami, S.O Skrovseth, and K.M. Augestad (2017). Analysis of Free Text in Electronic Health Records for Identification of Cancer Patient Trajectories. Nature Scientific Reports, doi:10.1038/srep462262017

A. Beskos, M. Girolami, S. Lan, P.E. Farrell, A.M. Stuart (2017). Geometric MCMC for Infinite-Dimensional Inverse Problems. Journal of Computational Physics, Volume 335, Pages 327–351,