自回归滑动平均建模的一种简便定阶方法 自回归滑动平均(n,n-1)建模策略
郭赟[1] 金水高[1] 刘丽华[2]
[1]中国疾病预防控制中心信息中心,北京100050 [2]解放军301总医院医疗统计科,北京100050
摘 要:
目的介绍时间序列中自回归滑动平均(ARMA)建模的一种简便定阶方法——ARMA(n,n-1)建模策略。方法简要说明ARMA(n,n-1)建模策略,并以模拟的ARMA(3,2)序列为例,进行ARMA(n,n-1)建模。结果结合AIC信息准则和残差序列图,确定了模型合理的阶数。结论ARMA(n,n-1)建模策略简便易懂,适合医学工作者采用。[著者文摘]
关 键 词:
文章出处:
《卫生研究》-2008年37卷1期 -102-103页
栏目信息:
分 类 号:
文献标识码:
A
文章编号:
1000-8020(2008)01-0102-02
A convenient way for autoregressive moving average model order determination autoregressive moving average ( n, n - 1) modeling strategy
GUO Yun, JIN Shuigao, LIU Lihua( Chinese Center for Disease Control and Prevention, Beijing 100050, China)
Abstract:
Objective To introduce the autoregressive moving average(ARMA) ( n, n - 1 ) modeling strategy as to a convenient way for the determination of ARMA orders, Methods ARMA ( n, n - 1 ) strategy was generally explained. A simulated ARMA (3,2) time series was generated and analized following this" strategy. Results Combined with AIC information criteria and residue plot, the order of ARMA model was determined and complied well with the simulated data. Conclusion ARMA( n, n - 1 ) strategy may be convenient for the medical workers.[著者文摘]
Key words:
atoregressive moving average model, autoregressive moving average( n, n - 1 ) modeling strategy

学术















cqvip.com