[1]邵 祥,杜秀丽.在ARFIMA模型中使用小波的另一种极大似然估计[J].南京师范大学学报(自然科学版),2010,33(01):16-21.
 Shao Xiang,Du Xiuli.Another Maximum Likelihood Estimator in an ARFIMA Model Using Wavelets[J].Journal of Nanjing Normal University(Natural Science Edition),2010,33(01):16-21.
点击复制

在ARFIMA模型中使用小波的另一种极大似然估计()
分享到:

《南京师范大学学报》(自然科学版)[ISSN:1001-4616/CN:32-1239/N]

卷:
第33卷
期数:
2010年01期
页码:
16-21
栏目:
数学
出版日期:
2010-03-20

文章信息/Info

Title:
Another Maximum Likelihood Estimator in an ARFIMA Model Using Wavelets
作者:
邵 祥1 杜秀丽2
1. 南京邮电大学通达学院, 江苏南京210003 2. 南京师范大学数学科学学院, 江苏南京210046
Author(s):
Shao Xiang1Du Xiuli2
1.College of Tongda,Nanjing University of Posts and Telecommunications,Nanjing 210003,China 2. School of Mathem atical Sciences, Nan jing Normal Univers ity, Nan jing 210046, China
关键词:
ARFIMA 模型 离散小波变换( DWT) 相合性 渐近正态性
Keywords:
m odel disc rete w avelet transform ( DWT) consistency asym ptotic normality
分类号:
O212.1
摘要:
用小波变换方法提出ARFIMA(p,d,q)模型中分形差分参数d的另一种极大似然估计(MLE).这种极大似然估计被证明是相合的和渐近正态的.仿真结果显示这种极大似然估计偏差较小,并且与Tse的估计值相比较有更小的根均方误差(RMSE),该方法估计长记忆时间序列是有效的.
Abstract:
Th is paper presents ano the rmax im um like lihood estim ator(MLE) o f the fractional differenc ing param eter d in an ARFIMA( p, d, q ) m ode l usingw ave lets transform. TheMLE is prov ed to be consist and asympto tic norm a .l The simu lation resu lts show th isM LE has low b ias and has low er root m ean square error( RMSE ) than the estima tor o fTse. Th is m ethod m ay be use fu l in estim ating long- memory time series.

参考文献/References:

[ 1] M a rk J Jensen. An alterna tive m ax imum like lihood estim ator o f long-m em ory pro cesses using compactly suppo rted w ave lets [ J] . Journa l of E conom ic Dynam ics Con tro ,l 2000, 24( 3): 361-387.
[ 2] Pe ter F Cra igm ile, Peter Guttorp, Dona ld B Perc iva.l W ave le t-based parame ter estim ation fo r po lynom ia l con tam inated fractionally d ifferenced processes[ J] . IEEE T ransactions on S ignal Processing, 2005, 53( 8): 3 151-3 161.
[ 3] T se Y K, Anh V V, Tieng Q. M ax im um like lihood estim ation o f the fractiona l d ifferenc ing param eter in an ARFIMA model using w ave lets[ J] . M athem atics and Computers in Sim ulation, 2002, 59( 1 /3): 153-161.
[ 4] H osking J R M. Frac tiona l d iffe rencing[ J]. B iom etr ika, 1981, 68( 1): 165-176.
[ 5] Brockwe ll P, Dav is R. T ime Ser ies: Theory andM ethods[M ]. 2nd ed. N ew York: Springer, 1991.
[ 6] Ba illieR T. Long-m em ory processes and fractiona l in teg ration in econome trics[ J] . Journa l o f Econom etr ics, 1996, 73( 1): 5-59.
[ 7] M cC oy E J, W a lden A T. W ave let ana ly sis and synthesis o f stationary long-memo ry pro cesses[ J] . Journa l o f Compu tational and G raph ica l Sta tistics, 1996, 5( 1): 26-56.
[ 8] Lehm ann E L. Theory of Po int Estima tion[M ]. N ew York: W iley, 1983.
[ 9] Donald B Perciva,l Andrew T W a lden. W aveletM ethods for Tim e Ser ies Analysis[M ]. Cambr idge: Cam bridge Un iversity Press, 2000.

备注/Memo

备注/Memo:
通讯联系人: 邵 祥, 助教, 研究方向: 时间序列分析. E-mail:tdshxi@njupt.edu.cn
更新日期/Last Update: 2013-04-08