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:Statistics
SUMMARY:Discrete Fourier transform methods in the analysis
of nonstationary time series - Suhasini Subba Rao
\, Texas A&\;M University
DTSTART;TZID=Europe/London:20130531T160000
DTEND;TZID=Europe/London:20130531T170000
UID:TALK45566AThttp://talks.cam.ac.uk
URL:http://talks.cam.ac.uk/talk/index/45566
DESCRIPTION:The Discrete Fourier Transform (DFT) is often used
to analysis time series\,\nusually under the assu
mption of (second order) stationarity of the time\
nseries. One of the main reasons for using this tr
ansformation is that the\nDFT tends to uncorrelate
the original time series. Thus the DFT can be\ntr
eated as an almost uncorrelated complex random var
iable\, and standard\nmethods for independent data
can be applied to the DFT. It can be shown that\n
this useful uncorrelation property does not hold f
or nonstationary time\nseries. This would suggest
that the DFT is no longer a helpful tool for\nnons
tationary time series analysis. However\, the purp
ose of this talk is to\ndemonstrate that correlati
ons between the DFTs contain useful information\na
bout the nonstationary nature of the underlying ti
me series. We will\nexploit the starkly contrastin
g correlation\nproperties between stationarity and
nonstationarity DFTs to construct a test\nfor sec
ond order stationarity and\, if time permits\, to
construct an\nestimator of the time-varying spectr
al density.\n
LOCATION:MR12\, Centre for Mathematical Sciences\, Wilberf
orce Road\, Cambridge
CONTACT:Richard Samworth
END:VEVENT
END:VCALENDAR