www.uhasselt.be
DSpace

Document Server@UHasselt >
Research >
Research publications >

Please use this identifier to cite or link to this item: http://hdl.handle.net/1942/281

Title: Kernel weighted influence
Authors: Hens, Niel
Aerts, Marc
Molenberghs, Geert
Thijs, Herbert
Verbeke, Geert
Keywords: Non and semiparametric methods
Missing data
Computer intensive
Issue Date: 2002
Citation: Stasinopoulos, Mikos & Touloumi, Giota (Ed.) Statistical Modelling. Proceedings of the 17th International Workshop on Statistical Modelling. Statistical Modelling in Society. p. 337-342.
Abstract: To asses the sensitivity for non-random dropout in a selection model framework, several methods were developed. None of them are without limitations. In this paper, a new method called kernel weighted influence is proposed. It uses several features of global and local influence approaches. Together with the use of nonparametric techniques, it provides a challenging new technique with a variety of options.
URI: http://hdl.handle.net/1942/281
Category: C2
Type: Proceedings Paper
Appears in Collections: Research publications

Files in This Item:

Description SizeFormat
N/A510.85 kBAdobe PDF

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.