http://swrc.ontoware.org/ontology#UnrefereedArticle
Misspecified MA (2) Model Fitting to a Data from Gaussian MA (q) Process
en
MA (q) model fitting
conditional maximum likelihood function
Gaussian
locally minimal points
misspecification.
Tanaka Minoru
This paper provides discussions about a misspecified MA (2) model fitting to (a data generated by) a Gaussian MA (q) process. They are mainly concerned a problem for finding a number of locally maximal points of the Gaussian quasi-maximum likelihood function of the model when the sample size is large. When 0 ≤ q ≤ 3, the MA (2) model has always only one parameter set estimated in the invertible parameter space. On the other hand when q ≥ 5, the likelihood function of the MA (2) model has more than one locally maximal points in the invertible parameter space if the model is fitted to a data from some MA (q) process.
The Institute of Information Science Senshu University
application/pdf
1349-1938
Information Science and Applied Mathematics
26
1-20
2018
|||