BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Simulation optimization via bootstrapped Kriging: survey - Kleijne
 n\, JPC (Universiteit van Tilburg)
DTSTART:20110907T110000Z
DTEND:20110907T113000Z
UID:TALK32693@talks.cam.ac.uk
CONTACT:Mustapha Amrani
DESCRIPTION:This presentation surveys simulation optimization via Kriging 
 (also called Gaussian Process or spatial correlation) metamodels. These me
 tamodels may be analyzed through bootstrapping\, which is a versatile stat
 istical method but must be adapted to the specific problem being analyzed.
  More precisely\, a random or discrete- event simulation may be run severa
 l times for the same scenario (combination of simulation input values)\; t
 he resulting replicated responses may be resampled with replacement\, whic
 h is called distribution-free bootstrapping. In engineering\, however\, de
 terministic simulation is often applied\; such a simulation is run only on
 ce for the same scenario\, so "parametric bootstrapping" is used. This boo
 tstrapping assumes a multivariate Gaussian distribution\, which is sampled
  after its parameters are estimated from the simulation input/output data.
  More specifically\, this talk covers the following recent approaches: (1)
  Efficient Global Optimiz ation (EGO) via Expected Improvement (EI) using 
 parametric bootstrapping to obtain an estimator of the Kriging predictor's
  variance accounting for the randomness resulting from estimating the Krig
 ing parameters. (2) Constrained optimization via Mathematical Programming 
 applied to Kriging metamodels using distribution-free bootstrapping to val
 idate these metamodels. (3) Robust optimization accounting for an environm
 ent that is not exactly known (so it is uncertain)\; this optimization may
  use Mathematical Programming and Kriging with distribution-free bootstrap
 ping to estimate the Pareto frontier. (4) Bootstrapped Kriging may preserv
 e a characteristic such as monotonicity of the outputs as a function of th
 e inputs.\n
LOCATION:Seminar Room 1\, Newton Institute
END:VEVENT
END:VCALENDAR