Statistics Group






The role of CSI in Cambridge Statistics
Some highlights
Collaborations
CSI organised meetings and other activities
Research Subjects

The role of CSI in Cambridge Statistics

In 2007 the Cambridge Statistics Initiative was established with a £2.3 million EPSRC Science & Innovation (S&I) Award, so creating a coherent research strategy for Statistics at Cambridge.

With the S&I award comes an institutional commitment to maintain a strong group in core and methodological Statistics within the Statistical Laboratory, and an that this group should lie at the heart of statistical activities in the University. A CSI lectureship based in Engineering has created effective links with researchers in signal processing, and there are many other substantial collaborations, including Astronomy, Oceanography and Archaeology. Fundamental research in causal inference has led to advances in medical and genetic statistics.

A central CSI focus, both because of its intrinsic importance and because of the special expertise of the group, is on developing new mathematical theory to address new statistical challenges. Nonparametric inference, high-dimensional variable selection and estimation problems are studied using sophisticated tools of mathematical analysis. This work is closely related to probability and approximation theory, and links with the newly created Cambridge Centre for Analysis. We shall promote the adoption of the resulting new statistical methods in scientific work at Cambridge and beyond.

 

Some highlights

  • Samworth awarded EPSRC Early Career Fellowship £ 1.2 million, 2012
  • Dawid will be giving the 30th Fisher memorial lecture in November 2011.
  • Spiegelhalter was appointed investigator on a £5 million grant to form the Cambridge Behaviour and Health Research Unit.
  • Samworth awarded Leverhulme Research Fellowship, 2011/12.
  • Samworth presented a read discussion paper on "Maximum likelihood estimation of a multi-dimensional log-concave density” at a Read Paper meeting of the Royal Statistical Society in 2010.
  • Samworth awarded the Guy medal in bronze from the Royal Statistical Society. The award, made in 2010, is with respect to his work on theoretical, methodological, and computational aspects of non-parametric statistics. The award is one of the society's most prestigious honours and is typically given to individuals under 35.
  • Samworth was awarded the biennial Royal Statistical Society Research prize for 2008.
  • Academy of Medical Sciences report `Identifying the environmental causes of disease: How should we decide what to believe and when to take action?', by a high-profile panel including Dawid.

Collaborations

Cambridge

  • Astrophysics/Rutherford Appleton Laboratory: Signal receptor networks
  • Department of Chemistry: Multi-object tracking
  • IBD Research Group, Division of Gastroenterology, Addenbrooke's Hospital: Disease causation
  • Institute of Criminology: Forecasting serious crime
  • MRC Biostatistics Unit, Cambridge: Genetic variation, Disease prediction and causation

UK

  • Bristol University: Causal inference
  • Bristol University/Imperial College London: Streaming data
  • Royal Holloway University of London: Game-theoretic probability

International

  • Humboldt University Berlin/Georgia Tech: Inverse Problems: Statistical Inference with Levy
    Processes
  • Max-Planck Institute for Biological Cybernetics, Tübingen: Robot imitation learning
  • McGill University: Bayesian clustering
  • Princeton University: Ultra-high dimensional feature selection
  • Princeton, North Carolina State and Columbia University: Iterative Sure Independence Screening
  • University of Berne: Log-concave distributions
  • University of Cagliari: Bayesian sensitivity analysis
  • University of Chicago Booth School of Business: Sequential Bayesian regularized regression
  • University of Connecticut/ENSAE Paris: Adaptive inference
  • ETH Zuerich, High Dimensional Statistical Inference
  • University of Paris VI & VII: Statistics on Manifolds, Nonparametric Bayesian Inference
  • University of Pavia, Department of Applied Health Sciences: Genetic disease prediction
  • University of Wisconsin: Text mining
  • University of Wollongong: Optimal bandwidth selection

Consultancy / Industrial Collaborations:

  • Aalborg Hospital, Denmark: High dimensional variable selection in cancer bioinformatics
  • http://www.askfood.gr: Bespoke recommender system
  • Atherosclerosis, Thrombosis and Vascular Biology (ATVB) Italian Investigators: Genetics of coronary artery disease
  • BAe Systems: Anomaly detection on graphs
  • British Antarctic Survey, Cambridge, UK: Modeling environmental phenomena
  • Broad Institute, Boston, U.S.A: Genetics of coronary artery disease
  • Centre for Research into Ecological and Environmental Modelling, St. Andrews: Ecological statistics
  • Genome Campus Sanger Center: Genetics of IBD
  • Qinetiq: Multi-object tracking
  • Met Office: Uncertainty in weather forecasting
  • NICE: Bayesian methodology
  • Spiral Software Ltd and Cambridge Temperature Concepts Ltd: Masters-levels projects
  • Wellcome Trust Sanger Institute, Hinxton, UK: Genetic disease prediction
  • Winton Capital: Public understanding of risk

    CSI organised meetings and other activities

    • As the host of the Academy for PhD Training in Statistics (APTS) in 01/2010 and 01/2011, offered statistics modules to 1st year PhD students from across the country.
    • Regular one-day meetings at which 50-60 Cambridge academics with statistical interests come together to present their work: this has stimulated contacts and collaborations in cancer research, criminology, machine learning, physics, systems biology, etc. The most recent meeting was held jointly with the Newton Institute programme ``Design of Experiments'' in September 2011.
    • Statistics Reading Group, attended by members of several departments.
    • ``Statistics Clinic'' (in collaboration with Physics and Engineering) to provide statistical advice and assistance to researchers across the University.
    • Two high-profile international conferences, ``Communicating Complex Statistical Evidence'', and ``Causal Inference: State of the Art''. Jointly organised a year-long Mellon-Sawyer Seminar Series ``Modelling Futures: Understanding Risk and Uncertainty''.
    • Several eminent foreign Statistics academics visited through a Kuwait Foundation scheme and Newton Institute programmes.


Research Subjects




Back to top