My Research Group contains members who are engaged in machine learning research, with a primary focus on Machine Learning Markets, probabilistic models for MRI, and sparse latent network models, with applications in epidemiology and finance. In addition, we are involved in time series problems (including continuous time systems), image analysis problems, non-parametric Bayesian methods, machine learning for music, and dataset shift and non-stationarity. The group is involved in both developing novel probabilistic and Bayesian methods, and employing them in these areas.
Colin Buchanan is working on analysis of tractography networks derived from diffusion MRI. | |
Jinli Hu is currently working on Machine Learning Markets, and alpha mixtures. | |
Zhanxing Zhu is working in the area of Machine Learning Markets. | |
Ksenia Andreyeva is working in the area of network analysis in diffusion MRI tractography. | |
Ben Pryke is looking at sparsity in deep learning. | |
Matthew Graham is working on the neural sampling hypothesis. | |
Gavin Gray is working on transactional filter graphs for machine learning. | |
Harri Edwards is working on adversarial methods for multimodal learning. |
Previous group members include
Tom Griffiths, who was working on methods for activity recognition, before moving to be involved in setting up companies Hubdub and FanDuel. | |
Ben Williams, who was involved in research on understanding motor primitives using handwriting motion as a structured basis for assessing the different components which go to make up a motion sequence. | |
Jon Clayden, who is developing methods for location and identification of fibre tracts from diffusion tensor imaging, and dealing with the issue of cross subject mapping and comparative analysis | |
Lawrence Murray whos research focused on particle filters for stochastic differential systems with application to fMRI. | |
James Withers who was working on obtaining high quality methods for registering EPI images to structural templates, and accurate brain region segmentation. | |
Bessi Bjarnason was researching methods of inference and learning in stochastic differential systems. | |
Felix Akagov was a research associate working on using machine learning methods in genetic epidemiology, specifically looking at methods for causal inference. | |
Andrew Dai was working on nonparametric methods for author disambiguation and analysis of conversational dynamics in meetings. | |
Athina Spiliopoulou wass involved in Machine Learning methods for music, with a specific interest in infering musical structure from melodic sequences. | |
David Reichert was examining the use of current machine learning models, such as Deep Boltzmann Machines, as appropriate models of neural compuation. | |
Krzysztof Gorgolewski was looking at the use of functional MRI and diffusion MRI methods for presurgical planning in neurosurgery. | |
Jono Millin was developing Market-Based mechanisms for machine learning, and went on to found the company behind DroneDeploy. | |
Jakub Piatkowski was working on diffusion MRI tractography and methods of infering tract width and tract density, which provides novel information from dMRI, and is a potentially useful biomarker in ageing. | |
Peter Orchard was working on methods for Portfolio management. | |
Simon Lyons was working on inference and learning in continuous-time systems. |