We are developing the next generation of AI technologies, such as deep reinforcement learning algorithms, which will provide machines with fully automated strategies for complex interaction tasks with the world. We are increasingly working with advanced simulations and “digital twins” to optimise physical systems and their control algorithms at the design stage, whether in the automotive industry or industrial robotics.

We are exploring the use of reinforcement learning and other AI technologies such as Bayesian Optimisation in computational settings rather than only physical systems. From our “automated neuroscientist” work to the potential automation of car crash simulations, we apply statistical machine learning solutions to important economic challenges.

We are developing computer vision systems to assist machines to extract information from static images and streaming videos, and are exploring applications in manufacturing, security and other industries. In Medical Image Processing we work with several NHS Trusts, the Wellcome Trust, and commercial customers to develop techniques for diagnosis and decision support in X-Ray imaging and other modalities such as MRI and CT. These have the potential to reduce or prioritise Radiologist’s workloads, support training, and achieve better patient outcomes.

Recent preprints can be found in arXiv. Below are some representative publications:

• Modelling imaging data
  
  • Tsagkrasoulis D and Montana G (2018) Random forest regression for manifold-valued responses. Pattern Recognition Letters
  • Cole J. et al. (2017) Predicting brain age with deep learning from raw imaging data results in a reliable and heritable biomarker. NeuroImage.
  • Tsagkrasoulis D et al. (2017) Heritability maps of human face morphology through large-scale automated three-dimensional phenotyping. Scientific Reports.
  • Ypsilantis P et al. (2015) Predicting response to neoadjuvant chemotherapy with PET imaging using convolutional neural networks. PloS ONE
  • de Brebisson A and Montana G (2015) Deep neural networks for anatomical brain segmentation. Computer Vision and Pattern Recognition Workshop
• Modeling network data
  
  • Monti R, Anagnostopoulos C and Montana G (2017) Learning population and subject-specific brain connectivity networks via mixed neighborhood selection. Annals of Applied Statistics
  • Monti R et al. (2017) Decoding time-varying functional connectivity networks via linear graph embedding methods. Frontiers of Neuroscience
  • Kuncheva Z and Montana G (2017) Multi-scale community detection in temporal networks using spectral graph wavelets. Workshop on Personal Analytics and Privacy. An Individual and Collective Perspective
  • Monti R et al. (2017) Real-time estimation of dynamic functional connectivity networks. Human Brain Mapping
  • Chung W et al. (2016) Characterising brain network topologies: a dynamic analysis approach using heat kernels. Neuroimage.
  • Kunceva Z and Montana G (2015) Community detection in multiplex networks using locally adaptive random walks. IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining
  • Ruan R, Young A and Montana G (2015). Differential analysis of biological networks. BMC Bioinformatics.
  • Monti R et al. (2014) Estimating time-varying brain connectivity networks from functional MRI time series. Neuroimage
• Automated experimental design with Bayesian optimisation
  • Lorenz R. et al. (2018) Dissociating frontoparietal brain networks with neuroadaptive Bayesian optimization. Nature Communications
  • Lorenz R. et al. (2016) The automatic neuroscientist: automated experimental design with real-time fMRI. Neuroimage
  • Lorenz R et al (2015) Stopping criteria for boosting automatic experimental design using real-time fMRI with Bayesian optimization. NIPS Workshop on Machine Learning and Interpretation in Neuroimaging: Beyond the Scanner
  • Lorenz R et al. (2016) Towards tailoring non-invasive brain stimulation using real-time fMRI and Bayesian optimization. International Workshop on Pattern Recognition in Neuroimaging
• Modelling genomic data - sparse regression methodologies
  
  • Krishnan M et al. (2018) Machine learning shows association between genetic variability in PPARG and cerebral connectivity in preterm infants. PNAS
  • Silver M et al. (2013) Pathways-driven sparse regression identifies pathways and genes associated with high-density lipoprotein cholesterol in two Asian cohorts. PLOS Genetics

  • Wang Z, Curry E. and Montana G (2014). Network-guided regression for detecting associations between DNA methylation and gene expression. Bioinformatics

  • Silver M and Montana G (2011) Fast identification of biological pathways associated with a quantitative trait using group lasso with overlaps. Statistical Applications in Genetics and Molecular Biology
  • Vounou M, Nichols T and Montana G (2010) Discovering genetic associations with high-dimensional neuroimaging phenotypes: a sparse reduced-rank regression approach. NeuroImage
• Modelling genomic data - distance-based methodologies
  
  • Minas C and Montana G (2014) Distance-based analysis of variance: approximate inference. Statistical Analysis and Data Mining
  • Minas C and Montana G (2013) A distance-based test of association between paired heterogeneous genomic data. Bioinformatics
  • Sim A, Tsagkrasoulis D, and Montana G (2013) Random forests on distance matrices for imaging genetics studies. Statistical Applications in Genetics and Molecular Biology

  • Wang et al. (2013) Random forests on Hadoop for genome-wide association studies of multivariate neuroimaging phenotypes. BMC Bioinformatics

  • Minas C, Waddell S and Montana G (2011) Distance-based differential analysis of gene curves. Bioinformatics
• Data clustering algorithms for high-dimensional data
  
  • McWilliams B and Montana G (2012) Multi-view predictive partitioning in high dimensions. Statistical Analysis and Data Mining
  • McWilliams B and Montana G (2014) Subspace clustering of high-dimensional data: a predictive approach. Knowledge Discovery and Data Mining
  • Cozzini A et al. (2014) A Bayesian mixture of lasso regressions with t-errors. Computational Statistics and Data Analysis
  • Cozzini A, Jasra A and Montana G (2013). Model-based clustering with gene ranking using penalised mixtures of heavy-tailed distributions. Journal of Bioinformatics and Computational Biology
• Modelling streaming data
  • Monti R, Anagnostopoulos C and Montana G (2018) Adaptive regularization for Lasso models in the context of non-stationary data streams. Statistical Analysis and Data Mining
  • McWilliams B and Montana G (2010) Sparse partial least squares for on-line variable selection in multivariate data streams. Statistical Analysis & Data Mining
  • Triantafyllopoulos K and Montana G (2011) Dynamic modeling of mean reverting spreads for statistical arbitrage. Computational Management Science
  • Montana G and Parrella F (2008) Learning to trade with incremental support vector regression experts. International Workshop on Hybrid Artificial Intelligence Systems
  • Montana G, Triantafyllopoulos K and Tsagaris T. (2008) Flexible least squares methods for temporal data mining and statistical arbitrage. Expert Systems and Applications

• Modelling text data
  

  • Cornegruta S. et al. (2016) Modelling radiological language with bidirectional long short-term memory networks. International Workshop on Health Text Mining and Information Analysis
  • Monti R. et al. (2016) Text-mining the NeuroSynth corpus using deep Boltzmann machines. International Workshop on Pattern Recognition in Neuroimaging