Generalized stochastic gradient learning

dc.creator	Evans, George W.
dc.creator	Honkapohja, Seppo
dc.creator	Williams, Noah
dc.date	2005
dc.date.accessioned	2013-10-16T07:02:31Z
dc.date.available	2013-10-16T07:02:31Z
dc.date.issued	2013-10-16
dc.identifier	http://hdl.handle.net/10419/19040
dc.identifier	ppn:503712469
dc.identifier.uri	http://koha.mediu.edu.my:8181/xmlui/handle/10419/19040
dc.description	We study the properties of generalized stochastic gradient (GSG) learning in forward-looking models. We examine how the conditions for stability of standard stochastic gradient (SG) learning both differ from and are related to E-stability, which governs stability under least squares learning. SG algorithms are sensitive to units of measurement and we show that there is a transformation of variables for which E-stability governs SG stability. GSG algorithms with constant gain have a deeper justification in terms of parameter drift, robustness and risk sensitivity.
dc.language	eng
dc.publisher
dc.relation	CESifo working papers 1576
dc.rights	http://www.econstor.eu/dspace/Nutzungsbedingungen
dc.subject	C65
dc.subject	C62
dc.subject	E17
dc.subject	E10
dc.subject	D83
dc.subject	ddc:330
dc.subject	adaptive learning
dc.subject	E-stability
dc.subject	recursive least squares
dc.subject	robust estimation
dc.subject	Rationale Erwartung
dc.subject	Lernprozess
dc.subject	Prognoseverfahren
dc.subject	Gleichgewichtsstabilität
dc.subject	Theorie
dc.title	Generalized stochastic gradient learning
dc.type	doc-type:workingPaper

الملفات في هذه المادة

الملفات	الحجم	الصيغة	عرض
لا توجد أي ملفات مرتبطة بهذه المادة.