Two-Parameter Ridge Estimation for the Coefficients of Almon Distributed Lag Model


ÖZBAY N.

IRANIAN JOURNAL OF SCIENCE AND TECHNOLOGY TRANSACTION A-SCIENCE, cilt.43, ss.1819-1828, 2019 (SCI İndekslerine Giren Dergi) identifier identifier

  • Cilt numarası: 43
  • Basım Tarihi: 2019
  • Doi Numarası: 10.1007/s40995-018-0634-5
  • Dergi Adı: IRANIAN JOURNAL OF SCIENCE AND TECHNOLOGY TRANSACTION A-SCIENCE
  • Sayfa Sayıları: ss.1819-1828

Özet

This paper resolves the difficulty in using Almon's technique in the estimation of distributed lag coefficients when the researchers run into the problem of multicollinearity. Multicollinearity arising in the analysis leads us to the biased estimation methods, one of which is Almon ridge estimator. In consideration of two-parameter ridge solution of Lipovetsky and Conklin (Appl Stoch Mod Bus Ind 21:525-540, 2005), we offer Almon two-parameter ridge estimator, which has advantage over Almon estimator as well as Almon ridge estimator with regard to mean square error criterion. Besides, an improvement in coefficient of multiple determination is observed for the new estimator relative to the Almon ridge estimator. An illustrative application of the mentioned estimators is examined by widely used Almon data to support the theoretical results. Additionally, a Monte Carlo experiment is performed to investigate the performance of the estimators in detail.