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:Adaptive design and control - Pronzato\, L (Univer
sit de Nice Sophia Antipolis)
DTSTART;TZID=Europe/London:20110720T090000
DTEND;TZID=Europe/London:20110720T100000
UID:TALK32105AThttp://talks.cam.ac.uk
URL:http://talks.cam.ac.uk/talk/index/32105
DESCRIPTION:There exist strong relations between experimental
design and control\, for instance in situations wh
ere optimal inputs are constructed in order to obt
ain precise parameter estimation in dynamical syst
ems or when suitably designed perturbations are in
troduced in adaptive control to force enough excit
ation into the system. The presentation will focus
on adaptive design when the construction of an op
timal experiment requires the knowledge of the mod
el parameters and current estimated values are sub
stituted for unknown true values. This adaptation
to estimated values creates dependency among obser
vations and makes the investigation of the asympto
tic behaviors of the design and estimator a much m
ore complicated issue than when the design is spec
ified independently of the observations. Also\, ev
en if the system considered is static\, this adapt
ation introduces some feedback and the adaptive-de
sign mechanism can be considered as a particular a
daptive-control scheme. The role of experimental d
esign in the asymptotic properties of estimators w
ill be emphasized. The assumption that the set of
experimental variables (design points) is finite f
acilitates the study of the asymptotic properties
of estimators (strong consistency and asymptotic n
ormality) in stochastic regression models. Two sit
uations will be considered: adaptive D-optimal des
ign and adaptive design with a cost constraint whe
re the design should make a compromise between max
imizing an information criterion (D-optimality) an
d minimizing a cost (function optimization). The c
ase when the weight given to cost minimization asy
mptotically dominates will be considered in detail
in connection with self-tuning regulation and sel
f-tuning optimization problems.\n
LOCATION:Seminar Room 1\, Newton Institute
CONTACT:Mustapha Amrani
END:VEVENT
END:VCALENDAR