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:Isaac Newton Institute Seminar Series
SUMMARY:Sequential detection of structural changes in irre
gularly observed data - Tobias Kley (Humboldt-Univ
ersität zu Berlin)
DTSTART;TZID=Europe/London:20180405T110000
DTEND;TZID=Europe/London:20180405T120000
UID:TALK103420AThttp://talks.cam.ac.uk
URL:http://talks.cam.ac.uk/talk/index/103420
DESCRIPTION:Online surveillance of time series is traditionall
y done with the aim to identify changes in the mar
ginal distribution under the assumption that the d
ata between change-points is stationary and that n
ew data is observed at constant frequency. In many
situations of interest to data analysts\, the cla
ssical approach is too restrictive to be used unmo
dified. We propose a unified system for the monito
ring of structural changes in streams of data wher
e we use generalised likelihood ratio-type statist
ics in the sequential testing problem\, obtaining
the flexibility to account for the various types o
f changes that are practically relevant (such as\,
for example\, changes in the trend of the mean).
The method is applicable to sequences where new ob
servations are allowed to arrive irregularly. Earl
y identification of changes in the trend of financ
ial data can assist to make trading more profitabl
y. In an empirical illustration we apply the proce
dure to intra-day prices of components of the NASD
AQ-100 stock market index. This project is joint w
ork with Piotr Fryzlewicz.
LOCATION:Seminar Room 2\, Newton Institute
CONTACT:INI IT
END:VEVENT
END:VCALENDAR