Department of Economics and Business Economics

A data-cleaning augmented Kalman filter for robust estimation of state space models

Research output: Contribution to journal/Conference contribution in journal/Contribution to newspaperJournal articleResearchpeer-review

Standard

A data-cleaning augmented Kalman filter for robust estimation of state space models. / Marczak, Martyna; Proietti, Tommaso; Grassi, Stefano.

In: Econometrics and Statistics, Vol. 5, No. 1, 01.01.2018, p. 107-123.

Research output: Contribution to journal/Conference contribution in journal/Contribution to newspaperJournal articleResearchpeer-review

Harvard

APA

CBE

MLA

Vancouver

Author

Marczak, Martyna ; Proietti, Tommaso ; Grassi, Stefano. / A data-cleaning augmented Kalman filter for robust estimation of state space models. In: Econometrics and Statistics. 2018 ; Vol. 5, No. 1. pp. 107-123.

Bibtex

@article{49373048817b484798ee3f72870ac8a2,
title = "A data-cleaning augmented Kalman filter for robust estimation of state space models",
abstract = "A robust augmented Kalman filter (AKF) is presented for the general state space model featuring non-stationary and regression effects. The robust filter shrinks the observations towards their one-step-ahead prediction based on the past, by bounding the effect of the information carried by a new observation according to an influence function. When maximum likelihood estimation is carried out on the replacement data, an M-type estimator is obtained. The performance of the robust AKF is investigated in two applications using as a modeling framework the basic structural time series model—a popular unobserved components model in the analysis of seasonal time series. First, a Monte Carlo experiment is conducted in order to evaluate the comparative accuracy of the proposed method for estimating the variance parameters. Second, the method is applied in a forecasting context to a large set of European trade statistics series.",
keywords = "Additive outlier, Augmented Kalman filter, DISTRIBUTIONS, Innovation outlier, OUTLIER DETECTION, Robust filtering, SCORING RULES, Structural time series model",
author = "Martyna Marczak and Tommaso Proietti and Stefano Grassi",
year = "2018",
month = "1",
day = "1",
doi = "10.1016/j.ecosta.2017.02.002",
language = "English",
volume = "5",
pages = "107--123",
journal = "Econometrics and Statistics",
issn = "2468-0389",
publisher = "Elsevier",
number = "1",

}

RIS

TY - JOUR

T1 - A data-cleaning augmented Kalman filter for robust estimation of state space models

AU - Marczak, Martyna

AU - Proietti, Tommaso

AU - Grassi, Stefano

PY - 2018/1/1

Y1 - 2018/1/1

N2 - A robust augmented Kalman filter (AKF) is presented for the general state space model featuring non-stationary and regression effects. The robust filter shrinks the observations towards their one-step-ahead prediction based on the past, by bounding the effect of the information carried by a new observation according to an influence function. When maximum likelihood estimation is carried out on the replacement data, an M-type estimator is obtained. The performance of the robust AKF is investigated in two applications using as a modeling framework the basic structural time series model—a popular unobserved components model in the analysis of seasonal time series. First, a Monte Carlo experiment is conducted in order to evaluate the comparative accuracy of the proposed method for estimating the variance parameters. Second, the method is applied in a forecasting context to a large set of European trade statistics series.

AB - A robust augmented Kalman filter (AKF) is presented for the general state space model featuring non-stationary and regression effects. The robust filter shrinks the observations towards their one-step-ahead prediction based on the past, by bounding the effect of the information carried by a new observation according to an influence function. When maximum likelihood estimation is carried out on the replacement data, an M-type estimator is obtained. The performance of the robust AKF is investigated in two applications using as a modeling framework the basic structural time series model—a popular unobserved components model in the analysis of seasonal time series. First, a Monte Carlo experiment is conducted in order to evaluate the comparative accuracy of the proposed method for estimating the variance parameters. Second, the method is applied in a forecasting context to a large set of European trade statistics series.

KW - Additive outlier

KW - Augmented Kalman filter

KW - DISTRIBUTIONS

KW - Innovation outlier

KW - OUTLIER DETECTION

KW - Robust filtering

KW - SCORING RULES

KW - Structural time series model

UR - http://www.scopus.com/inward/record.url?scp=85044462730&partnerID=8YFLogxK

U2 - 10.1016/j.ecosta.2017.02.002

DO - 10.1016/j.ecosta.2017.02.002

M3 - Journal article

VL - 5

SP - 107

EP - 123

JO - Econometrics and Statistics

JF - Econometrics and Statistics

SN - 2468-0389

IS - 1

ER -