Premium
Data Transformation and Self‐Exciting Threshold Autoregression
Author(s) -
Ghaddar D. K.,
Tong H.
Publication year - 1981
Publication title -
journal of the royal statistical society: series c (applied statistics)
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.205
H-Index - 72
eISSN - 1467-9876
pISSN - 0035-9254
DOI - 10.2307/2346347
Subject(s) - autoregressive model , transformation (genetics) , vector autoregression , econometrics , mathematics , chemistry , biochemistry , gene
S ummary We take the view that a time series model, linear or not, is judged adequate only if it reduces the observed data to approximate Gaussian white noise. We study the goodness of fit of self‐exciting threshold autoregressive models ( setar ) from this standpoint. We also study the practical utility of the instantaneous Box‐Cox transformation as an aid to facilitate the desired reduction. Multi‐step‐ahead predictions of the Wölfs sunspot numbers are given for the years 1980 to 1987.