Search

Copyright of Journal of the Institute of Science & Technology of Dumlupinar University / Dumlupinar Üniversitesi Fen Bilimleri Enstitüsü Dergisi is the property of Dumlupinar University, Institute of Science & Technology and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

In regression analysis, if there are some kind of relation (multicollinearity) between independent variables, the Least Squares estimation method may lead to the use of wrong models and hence to wrong findings out of the model. Various methods have been devised in order to carry out regression analysis with such independent variables which exhibit dependence on each other. One of such methods is the biased estimation methods. In this study, it has been studied on the necessary and sufficient conditions of biased estimators based on Ridge estimator to be at least efficient as much as the Least Squares estimator. [ABSTRACT FROM AUTHOR]

A new estimation concept for semiparametric regression models based on smoothing splines has been introduced in Eubank et al. (1998). The two different approaches based on smoothing splines have been compared by Schimek (2000) according to this estimation concept. In this study, the estimation of the semiparametric regression model is performed with two different approaches called partial splines of Green and Silverman and Speckman's approach and compared with the estimations of the ordinary least squares. The smoothing parameter is selected via generalized cross-validation in smoothing splines and inferences have been discussed for both the parametric and the nonparametric part of the model. For this purpose an application is performed by using a program coded in MATLAB. [ABSTRACT FROM AUTHOR]

In this paper, different regression models have been discussed for investigation of the relationships among house rents and trait variables in Eskişehir. According to statistical analysis, it is concluded that some of the explanatory variables have had linear and some of them have had nonparametric effect on house rents. Thus, the model has obtained as semiparametric adittive regression model that contain a few nonparametric components in additon to parametric linear components. It is observed that suitable semiparametric adittive regression model is significant, and given better results than parametric linear and semiparmetric models. [ABSTRACT FROM AUTHOR]

Copyright of Teknoloji is the property of Engineering Science & Technology, an International Journal and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)