COMBINING PATTERN-RECOGNITION TECHNIQUES WITH AKAIKES INFORMATION CRITERIA FOR IDENTIFYING ARMA MODELS

被引：13

作者：

WANG, L ^{[1
]}

LIBERT, GA ^{[1
]}

机构：

[1] FAC POLYTECH MONS,DEPT COMP SCI,MONS,BELGIUM

来源：

IEEE TRANSACTIONS ON SIGNAL PROCESSING | 1994年 / 42卷 / 06期

关键词：

D O I：

10.1109/78.286955

中图分类号：

TM [电工技术]; TN [电子技术、通信技术];

学科分类号：

0808 ; 0809 ;

摘要：

ARMA models are identified by combining pattern recognition techniques with Akaike's information criteria. First, pattern vectors of ARMA models are obtained by the extended sample autocorrelation functions method proposed by Tsay and Tiao [14]. Second, decision functions of various training samples are specified by the perceptron algorithm used in learning machines. Third, Akaike's AIC and BIC criteria are introduced. The practical utility of the proposed approach is illustrated by both simulated and practical data.

引用

页码：1388 / 1396

页数：9

共 16 条

[1]

AKAIKE H, 1979, BIOMETRIKA, V66, P237, DOI 10.1093/biomet/66.2.237

[2] NEW LOOK AT STATISTICAL-MODEL IDENTIFICATION [J].