Statistical Laboratory Seminars
Centre for Mathematical Sciences
Wilberforce Road, Cambridge, CB3 0WB
Tel: (01223) 337958
Fax: (01223) 337956
Email: [log in to unmask]
Michaelmas Term Seminars
Seminars will be held in Meeting Room 12, CMS
All interested are welcome
Friday 17 October
2.00pm Richard Smith (University of North Carolina, Chapel Hill)
Title: TBA
Friday 31 October
2.00pm Anthony Ledford (Man Investment Products)
The development of a trading system
In this talk an overview is given of the main issues that must be
addressed for the development of a quantitative trading system. Our
treatment focuses on futures markets as these are particularly suited to
systematic trading, and deals with modelling temporal dependence in both
mean and variance (volatility), trading rule selection, the effect of trading
costs, risk assessment and the benefits of diversification through
trading more than one market. These concepts will be illustrated using daily
data on the FTSE Allshare and S&P 500 indices and for simplicity we will
restrict attention to relatively straightforward statistical models.
Man Investment Products is the world's largest hedge fund manager and a
leading FTSE 100 company. We specialise in alternative investments and
manage funds through systematic trading strategies applied to a wide
portfolio of international markets. A diverse range of statistical
modelling and analysis techniques is exploited in researching and
developing these strategies.
Friday 14 November
2.00pm Joanne Waddington (MDS Pharma Services)
Careers for Statisticians within the Pharmaceutical Industry
Jo Waddington is a statistician working in the pharmaceutical industry.
The aim of this presentation is to raise awareness of the role of statistics
within the drug development arena and to encourage
students to consider a career as a statistician within the pharmaceutical
industry. The talk will describe the pharmaceutical industry, give an
overview of the drug development process and a detailed account of where a
statistician or statistical programmer might be involved. Career
opportunities and the
working environments within the industry are to be presented and key
skills and qualifications discussed. To conclude, an account of Jo's
career to date will be given, including an opportunity for the students to
ask questions. Various PSI brochures (including a careers booklet written
specifically for students considering
a career in statistics) and contact details for any future questions will
be provided. At the end of the talk there will be a prize draw for all
attendees with the winner receiving a £20 book token.
Friday 28 November
2.00pm Janet Heffernan (Lancaster University)
Extreme values in the dock: the formal enquiry into the sinking of the
MV Derbyshire
On the 9th September 1980 the bulk carrier MV-Derbyshire sank in the Pacific
when it was caught in a typhoon while transporting iron ore from Canada to
Japan. All 44 people on board died. The Derbyshire was a relatively new
ship, built to a high quality design by respected ship builders and run by
reputable company. It remains the largest UK ship lost at sea. The sinking
was unusual as the ship appears to have sunk without warning with no
may-day call being received and no lifeboats
launched.
In 1994 the wreckage was located 2.5 miles under the sea-surface and in 1996-7
the world's most expensive deep water survey provided considerable
information about the wreckage in the form of videos and photographs.
This led to the re-opening of the inquiry, the results of which were
heard in the summer of 2000. Early in the hearing it was decided that the
most likely cause of the ship's sinking was a sequence of events triggered
by a wave impact on the cover of the ship's front hold in excess of its
design strength. We became involved mid--way through the hearing when it
was realised that a statistical analysis of the extreme wave impacts on
the cover was required. As the court hearing was already underway, we were
asked to report within a very short time.
The statistical problem was to estimate the probability of the ship
receiving a wave impact on one of the hatch covers exceeding the design
strength of the cover. This estimated probability, with associated
confidence interval, was required for a range of hindcast sea-states and
possible ship conditions. Models motivated by
extreme value theory gave an excellent description of the large wave
impacts on the ship, covariate models accounted nicely for the effect of
explanatory variables, and likelihood-based inference proved to be an
ideal framework for modelling and reporting the statistical analysis.
The Judge's report was published in November 2000, clearing the crew of any
blame, and citing our contribution as ``of absolutely fundamental
importance to the outcome of this Investigation''. We have subsequently played
a substantial role in a large follow-up study which aimed to improve the
design standards of hold covers on bulk carriers.
(Joint work with Jonathan Tawn.)
Friday 5 December
2.00pm Caitlin Buck (Sheffield)
Statistical issues in constructing the new internationally agreed radiocarbon
calibration curve
In the early days of radiocarbon dating an assumption was made that the
proportion of carbon-14 in the earth's atmosphere has remained constant
over time and thus that there was a simple, linear relationship between
age as recorded by the decay of the radioactive isotope and calendar age.
It turns out, however, that there have been quite considerable and
non-monotonic fluctuations in the levels of carbon-14 in the earth's
atmosphere over time and so a calibration curve is needed to map from one
scale to the other.
Over the last two years an international team, consisting mostly of physicists
who work in radiocarbon dating laboratories, has been generating and
collating data suitable for aiding in the construction of a new
internationally agreed radiocarbon calibration curve. Such curves have
been constructed before, but this is the first time that the team has
included a statistician. In this seminar the statistician on the team will
talk about the nature of the data and the statistical and practical issues
involved in creating a calibration curve from them. Since there is
uncertainty in both the calendar ages and the radiocarbon ages of the
calibration data points, this is a non-standard calibration problem. The
approach that Dr Buck and a colleague (Paul Blackwell, Sheffield) have
taken is based on a Gaussian random walk model and produces a calibration
curve which differs somewhat in both nature and detail from those that
have gone before. Consequently, she will illustrate the old and new
calibration curves and highlight the impact of the changes on the user
communities concerned.
|
