2013-09-092013-06-102013-09-092013-04-172012-06-102013-06-03Joubert, M 2012, Nonparametric logistic regression using smoothing splines MSc dissertation, University of Pretoria, Pretoria, viewed yymmdd < http://upetd.up.ac.za/thesis/available/etd-06032013-114114 / >E13/4/501/gmhttp://hdl.handle.net/2263/30889Dissertation (MSc (Mathematical Statistics))--University of Pretoria, 2012.Logistic regression is a well-established technique for modeling a discrete response variable as a function of explanatory variables. The aim of this study is to introduce nonparametric regression using smoothing splines. Nonparametric regression yields a more flexible way of estimating curves and provides a larger set of functions to work from, which is not limited to a family of functions as in the parametric regression framework. Nonparametric regression falls into the framework of the general additive model with the natural cubic spline as the solution to the penalised least square criterion. Models are fitted using natural cubic splines. B-splines will be implemented due to their computational advantages gained from their almost orthogonal structure. Software, PROC IML within SAS, will be written to estimate these models.en© 2012 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria. No part of this work may be reproduced or transmitted in any form or by any means, without the prior written permission of the University of PretoriaUCTDNonparametric regressionSmoothing splinesDissertationhttp://upetd.up.ac.za/thesis/available/etd-06032013-114114/