2002-2003 Session
October 16th 2002 at MANDEC (Manchester Dental Education Centre), Higher Cambridge Street at 2pm Joint meeting Manchester University's Biostats Group on "Statistical analysis of images" Robert Aykroyd (Department of Statistics, University of Leeds) Adaptive Bayesian reconstruction of SPECT data. Statistical approaches to image reconstruction have been developed over the last 20 years and now produce generally good results. Although the approach has been widely accepted, there has been much discussion over precisely which models are most appropriate. In this talk the Bayesian approach will be reviewed and the use of Bayesian adaptive smoothing proposed as a general model for image processing applications. Chris Glasbey (Biomathematics and Statistics Scotland, JCMB) Analysis of microarrays and electrophoresis gels Micorarrays and gel electrophoresis are key technologies in genomics, and beyond. Statistical models will be presented for automating the identification and quantification of spots or bands for a range of problems. Methods take account of noise, spatial warping, image segmentation and use of reference datasets, and will be illustrated. Charles Taylor (Department of Statistics, University of Leeds) Muscle fibre images through point patterns and tesselations December 18th 2002 at MMU, 4.30pm for 5pm Statistical advice - would you pay for it? Larry Furlong, AstraZeneca Much of business can be seen as the management and improvement of processes, or making decisions with variable data. The statistical profession has the potential to make a major contribution to this broad area. However, our contribution is often underutilised, and we fail to gain the recognition we would like. A key question is how we value the contribution which we make, and convince others of this value. This talk will explore these issues, and learning, based on experiences within the Pharmaceutical industry. Part of the session will be interactive so that we can build on the audience's experiences. Feb 4th 2003 at Xaverian College, Rusholme at 2.30pm - note Tuesday! (held in the Drama Studio in Redcliffe Building) How to Win Games: Probability Models for Success David F. Percy, University of Salford Phil Taylor recently scored a nine-dart 501 finish in a world championship and Tony David made a 170 check-out in a similar tournament. What are the chances that you or I might achieve these magnificent feats? Peter Ebdon will soon defend his world snooker title under intense pressure from Stephen Hendry and others. How do they decide which shots to play and how could we improve our own games to emulate these masters? This light-hearted talk considers the beautiful application of probability models to everyday life, specifically focussing on popular games including those above and the national lottery, which has a huge impact in the UK. We also look at some puzzling paradoxes and some serious issues where probability is crucial, including a discussion of ongoing miscarriages of justice in society. (Dave is a senior lecturer at Salford, having gained a BSc in mathematics from Loughborough in 1985 and a PhD in statistics from Liverpool in 1990. His research and teaching interests are in Bayesian inference, multivariate analysis and stochastic processes, with applications to business, industry, medicine, law, games and defence. He has travelled extensively to international conferences, is a keen cricketer, enjoys hiking and sailing, and is currently appearing as second irate in the pantomime Peter Pan. He is also a committee member of the RSS Manchester Local Group) Directions: 2nd April 2003 at MMU, 4:30 for 5:00 (a very late March meeting!) Quality Issues in the 2001 Population Census Keith Whitfield, Office of National Statistics Key Statistics from the 2001 Census have recently been released, to be followed in the coming months by more detailed results. The processing of Census returns will be discussed showing what quality issues arose at each stage and how they were addressed. The main messages from the release of Key Statistics will also be considered. April 16th 2003 at MMU, 4.30pm for 5pm Whisky Classified - Seminar and Tutored Tasting David Wishart, University of St. Andrews The talk describes the cardinal flavours that can be found in single malt whiskies, and explains how the production and maturation processes develop these flavours. To simplify the comparison of malt whiskies, a standardised profile of 12 flavour features was developed from around a thousand tasting notes and over 500 whisky terms. The principal malt whiskies from all the Scottish distilleries were then profiled and grouped according to flavour. This helps answer questions like -
See also www.whiskyclassified.com/seminars.html Flyer
for this meeting in Word 97 (Secretary's note: This meeting deals with complex issues in cluster analysis. At the RSS Plymouth conference in 2002 it was one of the best attended meetings - even more so than some plenary talks. David tells me that the meeting is likely to go on rather past the ordinary finishing time) June 11th 2003 at MMU at 4.30pm for 5.00pm This meeting will include a very short AGM and report Graham Dunn (Biostats Group at Manchester University)The challenge of patient choice and non-adherence to treatment in RCTs of counselling and psychotherapy The talk will start with a short and relatively non-technical review of recent statistical ideas on the analysis and interpretation of results of RCTs in which there might be non-adherence to allocated treatment. This involves the use of potential outcomes (counterfactuals) in the definition and estimation of treatment effects. The ideas will be illustrated using data from the ODIN trial (BMJ 321, 1450-4, 2000). I will then move on to discuss how such methods might help us to evaluate the potential of partially- randomised 'patient preference' designs and to compare them with other, possibly more promising, alternatives. |
