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:Microsoft Research Cambridge\, public talks
SUMMARY:CMA-ES – a Stochastic Second-Order Method for Func
tion-Value FreeNumerical Optimization - Nikolaus H
ansen\, INRIA
DTSTART;TZID=Europe/London:20111109T140000
DTEND;TZID=Europe/London:20111109T150000
UID:TALK34378AThttp://talks.cam.ac.uk
URL:http://talks.cam.ac.uk/talk/index/34378
DESCRIPTION:We consider black-box optimization with little ass
umptions on the underlying objective function. Fur
ther\, we consider sampling from a distribution to
obtain new candidate solutions. Under mild assump
tions\, solving the original black-box optimizatio
n problem coincides with optimizing a parametrized
family of distributions of our choice. Choosing t
he family of multivariate normal distributions on
the continuous search domain\, a natural gradient
descent on this family leads to an instantiation o
f the so-called CMA-ES algorithm (covariance matri
x adaptation evolution strategy). In this talk\, t
he continuous black-box optimization problem will
be introduced and the CMA-ES algorithm will be ill
ustrated. The CMA-ES adapts a second-order model o
f the underlying objective function. On convex-qua
dratic functions\, the resulting covariance matrix
resembles the inverse Hessian matrix of the funct
ion. In contrast to Quasi-Netwon methods\, this ca
n be accomplished derivative- and even function-va
lue free. The CMA-ES reveals the same invariance p
roperties as the famous Nelder-Mead simplex downhi
ll method\, is robust and works reliably not only
in low dimensions and is surprisingly efficient on
convex as well as non-convex\, highly ill-conditi
oned problems.
LOCATION:Small lecture theatre\, Microsoft Research Ltd\, 7
J J Thomson Avenue (Off Madingley Road)\, Cambrid
ge
CONTACT:Microsoft Research Cambridge Talks Admins
END:VEVENT
END:VCALENDAR