BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//jEvents 2.0 for Joomla//EN
CALSCALE:GREGORIAN
METHOD:PUBLISH
BEGIN:VTIMEZONE
TZID:Europe/London
BEGIN:STANDARD
DTSTART:20190121T150000
RDATE:20190331T020000
TZOFFSETFROM:+0100
TZOFFSETTO:+0000
TZNAME:Europe/London GMT
END:STANDARD
BEGIN:STANDARD
DTSTART:20191027T010000
RDATE:20200329T020000
TZOFFSETFROM:+0100
TZOFFSETTO:+0000
TZNAME:Europe/London GMT
END:STANDARD
BEGIN:DAYLIGHT
DTSTART:20190331T020000
RDATE:20191027T010000
TZOFFSETFROM:+0000
TZOFFSETTO:+0100
TZNAME:Europe/London BST
END:DAYLIGHT
END:VTIMEZONE
BEGIN:VEVENT
UID:7cc57999bf95fac586b57c6004d41342
CATEGORIES:Featured events, RSS Webinar Series
CREATED:20191028T150240
SUMMARY:RSS Journal webinar
LOCATION:online
DESCRIPTION;ENCODING=QUOTED-PRINTABLE:\n\nPLEASE NOTE: This webinar was rescheduled from a previous date due to i
ndustrial action.\n\n\nThe RSS Journal webinar will now take place on 22&nb
sp;January at 3pm (UK time) and will feature two papers on online and
offline methods of change point detection methods:\n\n\nPaper 1: Onlin
e inference for multiple changepoint problems’ by Fearnhead & Liu.\nIt
was published in Series B, Volume 69, Issue 4, September 2007\nPresenter 1:
Paul Fearnhead.\n\n\nAbstract: We propose an on?line algorithm
for exact filtering of multiple changepoint problems. This algorithm enabl
es simulation from the true joint posterior distribution of the number and
position of the changepoints for a class of changepoint models. The computa
tional cost of this exact algorithm is quadratic in the number of observati
ons. We further show how resampling ideas from particle filters can be used
to reduce the computational cost to linear in the number of observations,
at the expense of introducing small errors, and we propose two new, optimum
resampling algorithms for this problem. One, a version of rejection contro
l, allows the particle filter to choose the number of particles that are re
quired at each time step automatically. The new resampling algorithms subst
antially outperform standard resampling algorithms on examples that we cons
ider; and we demonstrate how the resulting particle filter is practicable f
or segmentation of human G+C content.\n\n\nPaper 2: High dimensional c
hange point estimation via sparse projection’ by Wang & Samworth.\nIt w
as published in Series B, Volume 80, Issue1 in January 2018\nPresenter 2:&n
bsp; Tengyao Wang\n\n\nAbstract: Change points are a very common feat
ure of ‘big data’ that arrive in the form of a data stream. We study high d
imensional time series in which, at certain time points, the mean structure
changes in a sparse subset of the co?ordinates. The challenge is to borrow
strength across the co?ordinates to detect smaller changes than could be o
bserved in any individual component series. We propose a two?stage procedur
e called inspect for estimation of the change points: first, we argue that
a good projection direction can be obtained as the leading left singular ve
ctor of the matrix that solves a convex optimization problem derived from t
he cumulative sum transformation of the time series. We then apply an exist
ing univariate change point estimation algorithm to the projected series. O
ur theory provides strong guarantees on both the number of estimated change
points and the rates of convergence of their locations, and our numerical
studies validate its highly competitive empirical performance for a wide ra
nge of data?generating mechanisms. Software implementing the methodology is
available in the R package InspectChangepoint.\n\n\nChair: Yi Yu, Universi
ty of Warwick\nDiscussant: Claudia Kirch, Otto von Guericke University Magd
eburg\n\n\nThe Wang & Samworth paper is open access. The Fearnhead &
; Liu paper will be available to download free of charge from two weeks bef
ore the webinar.\n\n\nFurther details are available at rss.org.uk/journal-w
ebinar\n
CONTACT:Judith Shorten
DTSTAMP:20200228T034946Z
DTSTART;TZID=Europe/London:20200122T150000
SEQUENCE:0
TRANSP:OPAQUE
END:VEVENT
END:VCALENDAR