BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:System identification: General aspects and structure.
DTSTART:20091127T101500
DTSTAMP:20260407T230322Z
UID:cad7fa4b85d1f2657477713c96b52652ab6fc4abe9792612b8733a11
CATEGORIES:Conferences - Seminars
DESCRIPTION:Prof. M. Deistler\, Institute for Mathematical Methods in Econ
 omics (Vienna University of Technology). \nSystem identification is concer
 ned with obtaining good models from data\, i.e. with data-\ndriven modelin
 g. In this contribution\, the aim is to explain and discuss ideas\, genera
 l \napproaches\, theories and algorithms for the identification of multi-i
 nput\, multi-output linear \ndynamic systems with stochastic noise. Identi
 fication of linear systems is a nonlinear \nproblem\, since the function a
 ttaching an estimated system to the data is nonlinear and is \nprototypica
 l also for many parts of identification of nonlinear systems. We discuss p
 roblems \nof structure theory\, such as parametrization\, estimation of re
 al-valued parameters\, \n(maximum likelihood type estimation and its asymp
 totic properties) and model selection \n(order estimation by information c
 riteria).
LOCATION:MEC2405 
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR