Premium
A remark on least‐squares and naive extrapolations in non‐linear AR (1) processes
Author(s) -
Anděl Jiří
Publication year - 1996
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/(sici)1099-131x(199612)15:7<549::aid-for630>3.0.co;2-f
Subject(s) - extrapolation , mathematics , least squares function approximation , class (philosophy) , generalized least squares , autoregressive model , linear model , statistics , computer science , artificial intelligence , estimator
The m ‐step least squares extrapolation is generally different from the m ‐step naive extrapolation in non‐linear AR (1) models when m ≥ 2. We show that there exists a class of non‐linear AR (1) models in which a difference between these two extrapolations is arbitrary large.