Comparison of some penalized methods in time series models
Tahir R. Dikheel Alaa Q. Yaseen
Department of Statistics,
College of Administration and Economics, University of Al-Qadisiyah, Iraq.
| Corresponding Author: Tahir R. Dikheel Alaa Q. Yaseen |
| Abstract : In this paper , the lag weighted lasso (lwlasso) is compared with lasso and alasso to deal with effect of lag in linear time series models. The lwlasso methods are more stable than the other methods in the comparison. Consequently, lag weighted lasso methods are capable to dealing with lag effect . In particular, the lag weighted lasso methods with w^R2 and w^R3 weights gave the best results compared with the other methods. |
