Geodesic
Exponential smoothing for time series with outliers
Kybernetika, Tome 47 (2011) no. 2, pp. 165-178
Cet article a éte moissonné depuis la source Czech Digital Mathematics Library

Voir la notice de l'article

Recursive time series methods are very popular due to their numerical simplicity. Their theoretical background is usually based on Kalman filtering in state space models (mostly in dynamic linear systems). However, in time series practice one must face frequently to outlying values (outliers), which require applying special methods of robust statistics. In the paper a simple robustification of Kalman filter is suggested using a simple truncation of the recursive residuals. Then this concept is applied mainly to various types of exponential smoothing (recursive estimation in Box-Jenkins models with outliers is also mentioned). The methods are demonstrated using simulated data.
Recursive time series methods are very popular due to their numerical simplicity. Their theoretical background is usually based on Kalman filtering in state space models (mostly in dynamic linear systems). However, in time series practice one must face frequently to outlying values (outliers), which require applying special methods of robust statistics. In the paper a simple robustification of Kalman filter is suggested using a simple truncation of the recursive residuals. Then this concept is applied mainly to various types of exponential smoothing (recursive estimation in Box-Jenkins models with outliers is also mentioned). The methods are demonstrated using simulated data.
Classification : 60G35, 62M10, 62M20, 90A20
Keywords: exponential smoothing; Kalman filter; outliers; robust smoothing and forecasting 
@article{KYB_2011_47_2_a0,
     author = {Hanz\'ak, Tom\'a\v{s} and Cipra, Tom\'a\v{s}},
     title = {Exponential smoothing for time series with outliers},
     journal = {Kybernetika},
     pages = {165--178},
     year = {2011},
     volume = {47},
     number = {2},
     mrnumber = {2828571},
     zbl = {1220.62114},
     language = {en},
     url = {http://geodesic.mathdoc.fr/item/KYB_2011_47_2_a0/}
}
TY  - JOUR
AU  - Hanzák, Tomáš
AU  - Cipra, Tomáš
TI  - Exponential smoothing for time series with outliers
JO  - Kybernetika
PY  - 2011
SP  - 165
EP  - 178
VL  - 47
IS  - 2
UR  - http://geodesic.mathdoc.fr/item/KYB_2011_47_2_a0/
LA  - en
ID  - KYB_2011_47_2_a0
ER  - 
%0 Journal Article
%A Hanzák, Tomáš
%A Cipra, Tomáš
%T Exponential smoothing for time series with outliers
%J Kybernetika
%D 2011
%P 165-178
%V 47
%N 2
%U http://geodesic.mathdoc.fr/item/KYB_2011_47_2_a0/
%G en
%F KYB_2011_47_2_a0
Hanzák, Tomáš; Cipra, Tomáš. Exponential smoothing for time series with outliers. Kybernetika, Tome 47 (2011) no. 2, pp. 165-178. http://geodesic.mathdoc.fr/item/KYB_2011_47_2_a0/

[1] Abraham, B., Ledolter, J.: Statistical Methods for Forecasting. Wiley, New York 1983. | MR | Zbl

[2] Anděl, J., Zichová, J.: A method for estimating parameter in nonnegative MA(1) models. Comm. Statist. Theory Methods 31 (2002), 2101–2111. | DOI | MR | Zbl

[3] Cipra, T.: Some problems of exponential smoothing. Appl. Math. 34 (1989), 161–169. | MR | Zbl

[4] Cipra, T.: Robust exponential smoothing. J. Forecasting 11 (1992), 57–69. | DOI

[5] Cipra, T., Romera, R.: Robust Kalman filter and its applications in time series analysis. Kybernetika 27 (1991), 481–494. | MR

[6] Cipra, T., Trujillo, J., Rubio, A.: Holt-Winters method with missing observations. Management Sci. 41 (1995), 174–178. | DOI | Zbl

[7] Croux, C., Gelper, S., Fried, R.: Computational aspects of robust Holt-Winters smoothing based on M-estimation. Appl. Math. 53 (2008), 163–176. | DOI | MR | Zbl

[8] Gardner, E. S.: Exponential smoothing: The state of the art. J. Forecasting 4 (1985), 1–28.

[9] Gardner, E. S.: Exponential smoothing: The state of the art - Part II. Internat. J. Forecasting 22 (2006), 637–666. | DOI

[10] Gelper, S., Fried, R., Croux, C.: Robust forecasting with exponential and Holt-Winters smoothing. J. Forecasting 29 (2010), 285–300. | MR | Zbl

[11] Taylor, J. W.: Smooth transition exponential smoothing. J. Forecasting 23 (2004), 385–404. | DOI

[12] Yohai, V., Zamar, R.: High break down point estimates of regression by means of the minimization of an efficient scale. J. Amer. Statist. Assoc. 83 (1988), 406–413. | DOI | MR