BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//talks.cam.ac.uk//v3//EN
BEGIN:VTIMEZONE
TZID:Europe/London
BEGIN:DAYLIGHT
TZOFFSETFROM:+0000
TZOFFSETTO:+0100
TZNAME:BST
DTSTART:19700329T010000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=-1SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0100
TZOFFSETTO:+0000
TZNAME:GMT
DTSTART:19701025T020000
RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=-1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
CATEGORIES:Kuwait Foundation Lectures
SUMMARY:Technology-driven statistics - Professor Terry Spe
ed (Berkeley)
DTSTART;TZID=Europe/London:20080219T170000
DTEND;TZID=Europe/London:20080219T180000
UID:TALK8787AThttp://talks.cam.ac.uk
URL:http://talks.cam.ac.uk/talk/index/8787
DESCRIPTION:Once upon a time most statistical inference was ca
rried out by regarding the data being analyzed as
realizations of random variables whose joint distr
ibution was determined up to some unknown paramete
rs\, usually but not always finite-dimensional. T
he challenge lay in dealing with the unknowns when
making appropriate inferences. This view enabled
us to draw on a fine body mathematical theory\, w
hich was comforting in that there seemed to be a s
olid foundation for what we were doing\, going bac
k to Kolmogorov's 1933 axiomatization of probabili
ty. (Of course this is a gross oversimplification
\, and ignores major philosophical issues.) In h
is famous 1962 paper "The future of data analysis"
Tukey questioned this orthodoxy\, and promoted da
ta analysis\, a subject related to statistics\, bu
t one far less governed by mathematical theory\, a
nd which did not appear to have any foundations.
Tukey's view is flourishing today\, yet mathematic
al statistics lives\, and may itself be flourishin
g. We now have many processes - assays\, devices\
, technologies - which can generate large amounts
of data very quickly\, data for which a realistic
joint distribution is unimaginable\, no matter how
we might parameterize. By this I simply mean tha
t we could never pass the statistician's Turing te
st - to simulate data indistinguishable from the r
eal thing - with such data. What do we do? Well\
, we combine statistics with data analysis (as per
haps we always have)\, doing things that seem appr
opriate\, alongside with things that would be corr
ect\, given certain assumptions that are patently
false. I'll be illustrating these ideas with exa
mples from biology\, more precisely\, high-through
put biology. It often seems to work\, sometimes
rather well\, and one day we may understand why.
LOCATION:Wolfson Room (MR 2) Centre for Mathematical Scienc
es\, Wilberforce Road\, Cambridge
CONTACT:Helen Innes
END:VEVENT
END:VCALENDAR