Bayesian Analysis of High Dimensional Data
Registration is now closed.
April 14th-16th, 2008
The primary aim of the workshop is to highlight recent methodological and applied advances in the Bayesian analysis of complex data.
The last decade witnessed an unprecedented growth in the generation of large amounts of heterogeneous datasets. Typical examples are satellite images and environmental measurements, fingerprints, gene and protein expression, astronomical data, high frequency financial time series, brain imaging and multiple spike trains recordings. The substantial experimental error affecting these data requires developing appropriate statistical methodologies for separating signal and noise. The fast-paced accumulation of such complex data underscores the key role of Bayesian methods for synthesizing the information accrued and for out-of-sample predictions.
The workshop will be organised into five keynote presentations and several related invited presentations. Confirmed speakers are:
Nils Hjort (University of Oslo)
Rob Kass (Carnegie Mellon University, USA)
Douglas Nychka (National Center for Atmospheric Research, USA)
Carl Rasmussen (University of Cambridge, UK)
David Madigan (Columbia University, USA)
Malay Ghosh (University of Florida, USA)
Michele Guindani (University of New Mexico, USA)
John Haslett (Trinity College Dublin)
Feng Liang (Duke University, USA)
Jeffrey Morris (MD Anderson Cancer Center, USA)
Bruno Sanso' (University of California at Santa Cruz, USA)
Yee Whye Teh (Gatsby Computational Neuroscience Unit, UK)
Darren Wilkinson (University of Newcastle, UK)
In addition to the oral presentations, a joint poster session with the parallel workshop on Composite Likelihood Methods will provide younger researchers and research students with an important opportunity to present their work.
This workshop is a satellite event within the program “Statistical Theory and Methods for Complex, High Dimensional Data” of the Isaac Newton Institute for Mathematical Sciences.
The meeting organisers are:
PARALLEL CRiSM WORKSHOPS:
Participants are encouraged also to take part to the CRiSM-organised workshop on Composite Likelihood Methods. The two workshops will also combine to organise a single poster session and social event, in order to maximize the impact of contributed posters. Details of the registration process for both workshops will appear here soon.