Courses
We offer courses in semiparametric modelling/generalized additive models, in group sequential and adaptive clinical trials and Bayesian hierarchical modelling.
Bayesian hierarchical modelling and WinBugs
Gavin Shaddick gives courses on Bayesian modelling. These courses range from half a day (Introduction to WinBugs) to two days (Bayesian hierarchical modelling).
Some example course material is available at web page. Email g.shaddick@bath.ac.uk if you are interested in arranging a course.
GAMs and other penalized GLMs in R
Simon Wood is the author of the R recommended semiparametric modelling package mgcv available from CRAN and shipped with all R distributions. Simon has given courses on smooth modelling with mgcv, ranging from half a day to a week in duration, to audiences including academics, government researchers and industrial statisticians. Recent courses have been in Tampere, Finland (Finnish Biostatistics society); University of Gronningen, Netherlands; UseR Conference, Rennes, France; University of Pamplona, Spain; CSIRO, Tasmania and IFREMAR, Nante France.
Some example course material is available at Simon's mgcv page. Email s*wood@bath*ac*uk (where '*' is really '.') if you are interested in arranging a course.
Group sequential and adaptive clinical trials
Chris Jennison is co-author, with Bruce Turnbull, of the book 'Group Sequential Methods with Applications to Clinical Trials'. He has given courses on group sequential and adaptive clinical trials ranging from half a day to two days in duration, to audiences including clinical biostatisticians and statisticians in pharmaceutical industry; recent courses have been at the Pharma IQ Conference on Innovation in Clinical Design (London, 2010); VIB Adaptive Trials Conference (Brussels, 2009); International Society for Clinical Biostatistics Congress (Copenhagen, 2008); Pfizer UK (Sandwich, 2007); Deming Conference (Atlantic City, 2006).
Email c.jennison@bath.ac.uk if you are interested in arranging a course.
Consultancy
Many members of the group engage in statistical consultancy. Example projects are given below.- Improved statistical inference from gas turbine rim sealing experiments. This project with engineers at Bath provided improved methods for estimating the parameters of nonlinear theoretical rim sealing models from experimental rig data (Simon Wood.)
- Improved short term prediction of the load on the French electricity grid. This project with workers at EDF developed methods to improve short term load prediction model using semi-parametric models, by developing fitting and update algorithms capable of handling very large datasets with serial autocorrelation. (Simon Wood.)
- Improved trend estimation for tree health in Baden-Wuerttemberg and optimised monitoring sampling scheme under given financial constraints. (Nicole Augustin.)
- Improved predictions of wind speeds for sites where no wind speed data is available. In this project feasibility assessment of prospective installations of wind power systems were performed using Bayesian modelling. (Gavin Shaddick.)
- Improved methodology of statistical models used for predicting capital maintenance costs in the water industry. (Gavin Shaddick.)
- Study design for assessing future customer service requirements in the water industry. (Gavin Shaddick.)
- Methodology for classifying customer groups according to energy usage. (Gavin Shaddick.)
- Study design and analysis of assessment of information campaign on avoidance of childhood accidents. (Gavin Shaddick.)
- Improved automotive parts for increased robustness in the face of environmental variation. (Julian Faraway)
- Improved product distribution of vehicle models for increased sales. (Julian Faraway)
- Improved screening processes for chemical discovery. (Julian Faraway)
- Tested for consumer product effectiveness. (Julian Faraway)
- Consultancy on clinical trial design. (Chris Jennison)