Estimation in semi-parametric and additive regression using smoothing and regression spline

Aydin D., TÜZEMEN M. Ş.

2nd International Conference on Computer Research and Development, ICCRD 2010, Kuala-Lumpur, Malaysia, 7 - 10 May 2010, pp.465-469 identifier

  • Publication Type: Conference Paper / Full Text
  • Doi Number: 10.1109/iccrd.2010.101
  • City: Kuala-Lumpur
  • Country: Malaysia
  • Page Numbers: pp.465-469
  • Keywords: Additive model, Penalized least squares, Regression spline, Semi-parametric model, Smoothing spline
  • Anadolu University Affiliated: Yes


In this paper, we try to investigate how the current account to gross domestic product (CAGDP) ratio in Turkey reacts to variations in the real effective exchange rate (REER) over the period of 1991:Q4-2007:Q3. For this purpose, we consider two different nonparametric regression models. In both approaches the TIME variable is specified as nonparametrically, but the REER variable is different. Firstly, we discuss a semi-parametric regression model where the parametric part is REER variable. Secondly, we consider an additive model where one of the nonparametric parts is a smooth function of REER variable and, therefore, the model falls within the class of additive models. The obtained findings showed that the response of the CAGDP ratio is negative nonlinear function of REER. Further, our estimations demonstrate that the coefficient giving the response of the CAGDP ratio to a change in REER declines over time when we assume a semi-parametric model. © 2010 IEEE.