Premium
POWER OF THE NEURAL NETWORK LINEARITY TEST
Author(s) -
Teräsvirta Timo,
Lin ChienFu,
Granger Clive W. J.
Publication year - 1993
Publication title -
journal of time series analysis
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.576
H-Index - 54
eISSN - 1467-9892
pISSN - 0143-9782
DOI - 10.1111/j.1467-9892.1993.tb00139.x
Subject(s) - mathematics , artificial neural network , linearity , lagrange multiplier , multiplier (economics) , series (stratigraphy) , simple (philosophy) , test (biology) , power (physics) , score test , algorithm , mathematical optimization , statistical hypothesis testing , statistics , artificial intelligence , computer science , electronic engineering , paleontology , philosophy , physics , epistemology , quantum mechanics , biology , engineering , economics , macroeconomics
. Recently, a new linearity test for time series was introduced based on concepts from the theory of neural networks. Lee et al. have already studied the power properties of this test and they are further investigated here. They are compared by simulation with those of a Lagrange multiplier (LM) type test that we derive from the same single‐hidden‐layer neural network model. The auxiliary regression of our LM type test is a simple cubic ‘dual’ of the Volterra expansion of the original series, and the power of the test appears superior overall to that of the other test.