Bandwidth selection for local polynomial smoothing of multinomial data

Please use this identifier to cite or link to this item: http://hdl.handle.net/1942/3143

Title:	Bandwidth selection for local polynomial smoothing of multinomial data
Authors:	Augustyns, Ilse Wand, MP
Issue Date:	1998
Publisher:	PHYSICA VERLAG GMBH
Source:	COMPUTATIONAL STATISTICS, 13(4). p. 447-461
Abstract:	We develop a rule for choosing bandwidths for local polynomial smoothing of ordered multinomial data. Our method is a variant of the double smoothing idea and is particularly geared towards good performance near the boundaries of the data, through the use of exact risk expressions.
Notes:	Limburgs Univ Ctr, Dept Stat, B-3590 Diepenbeek, Belgium. Univ New S Wales, Australian Grad Sch Management, Sydney, NSW 2052, Australia.Augustyns, I, Limburgs Univ Ctr, Dept Stat, Univ Campus, B-3590 Diepenbeek, Belgium.
Keywords:	categorical data; contingency table; double smoothing; Kernel estimator; sparse multinomial data
Document URI:	http://hdl.handle.net/1942/3143
ISSN:	0943-4062
e-ISSN:	1613-9658
ISI #:	000077902300002
Type:	Journal Contribution
Validations:	ecoom 2000
Appears in Collections:	Research publications

Check