Premium
Identification of TAR models using recursive estimation
Author(s) -
Bermejo Miguel Ángel,
Peña Daniel,
Sánchez Ismael
Publication year - 2011
Publication title -
journal of forecasting
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.543
H-Index - 59
eISSN - 1099-131X
pISSN - 0277-6693
DOI - 10.1002/for.1188
Subject(s) - autoregressive model , identification (biology) , computer science , monte carlo method , tar (computing) , estimation , econometrics , algorithm , statistics , mathematics , engineering , botany , systems engineering , biology , programming language
This paper proposes an automatic procedure to identify threshold autoregressive models and specify the values of thresholds. The proposed procedure is based on the time‐varying estimation of the parameters using an arranged autoregression. The proposed method not only allows for the automatic identification of the thresholds, but also has a superior identification performance than the competitors. The performance of the proposed procedure is illustrated using Monte Carlo experiments and real data. Copyright © 2010 John Wiley & Sons, Ltd.