Title :
Criterion for selecting model order in identification
Author :
Kwon, Soon H. ; Sugeno, M.
Author_Institution :
Sch. of E&E, Yeungnam Univ., Kyongbun, South Korea
fDate :
2/18/1999 12:00:00 AM
Abstract :
The authors propose a consistent and bias corrected extension of Akaike´s information criterion (AIC), AICbc. Empirical performances of the AICbc are studied. the AICbc was found to result in relatively better model order choices of a linear regression model using a Monte Carlo experiment
Keywords :
Monte Carlo methods; identification; statistical analysis; Monte Carlo experiment; bias corrected extension; information criterion; linear regression model; model order selection; selection criterion; system identification;
Journal_Title :
Electronics Letters
DOI :
10.1049/el:19990201
