Home > Research > Publications & Outputs > Robust estimation methods for a class of log-li...

Research

Associated organisational unit

Links

Text available via DOI:

Keywords

View graph of relations

Robust estimation methods for a class of log-linear count time series models

Research output: Contribution to Journal/Magazine › Journal article › peer-review

Published

Standard

Robust estimation methods for a class of log-linear count time series models. / Kitromilidou, S.; Fokianos, K.
In: Journal of Statistical Computation and Simulation, Vol. 86, No. 4, 2016, p. 740-755.

Research output: Contribution to Journal/Magazine › Journal article › peer-review

Harvard

Kitromilidou, S & Fokianos, K 2016, 'Robust estimation methods for a class of log-linear count time series models', Journal of Statistical Computation and Simulation, vol. 86, no. 4, pp. 740-755. https://doi.org/10.1080/00949655.2015.1035271

APA

Kitromilidou, S., & Fokianos, K. (2016). Robust estimation methods for a class of log-linear count time series models. Journal of Statistical Computation and Simulation, 86(4), 740-755. https://doi.org/10.1080/00949655.2015.1035271

Vancouver

Kitromilidou S, Fokianos K. Robust estimation methods for a class of log-linear count time series models. Journal of Statistical Computation and Simulation. 2016;86(4):740-755. Epub 2015 Apr 21. doi: 10.1080/00949655.2015.1035271

Author

Kitromilidou, S. ; Fokianos, K. / Robust estimation methods for a class of log-linear count time series models. In: Journal of Statistical Computation and Simulation. 2016 ; Vol. 86, No. 4. pp. 740-755.

Bibtex

@article{269f48defc00443986eab75a8059083d,
title = "Robust estimation methods for a class of log-linear count time series models",
abstract = "We study robust estimation of a log-linear Poisson model for count time series analysis. More specifically, we study robust versions of maximum likelihood estimators (MLEs) under three different forms of interventions: additive outliers (AOs), transient shifts (TSs) and level shifts (LSs). We estimate the parameters using the MLE, the conditionally unbiased bounded-influence estimator and the Mallows quasi-likelihood estimator and compare all three estimators in terms of their mean-square error, bias and mean absolute error. Our empirical results illustrate that under a LS or a TS there are no significant differences among the three estimators and the most interesting results are obtained in the presence of AOs. The results are complemented by a real data example.",
keywords = "autocorrelation, canonical link, conditionally unbiased bounded-influence estimator, interventions, log-linear Poisson model, Mallows quasi-likelihood estimator, tuning constant, simulation",
author = "S. Kitromilidou and K. Fokianos",
year = "2016",
doi = "10.1080/00949655.2015.1035271",
language = "English",
volume = "86",
pages = "740--755",
journal = "Journal of Statistical Computation and Simulation",
issn = "0094-9655",
publisher = "Taylor and Francis Ltd.",
number = "4",

}

RIS

TY - JOUR

T1 - Robust estimation methods for a class of log-linear count time series models

AU - Kitromilidou, S.

AU - Fokianos, K.

PY - 2016

Y1 - 2016

N2 - We study robust estimation of a log-linear Poisson model for count time series analysis. More specifically, we study robust versions of maximum likelihood estimators (MLEs) under three different forms of interventions: additive outliers (AOs), transient shifts (TSs) and level shifts (LSs). We estimate the parameters using the MLE, the conditionally unbiased bounded-influence estimator and the Mallows quasi-likelihood estimator and compare all three estimators in terms of their mean-square error, bias and mean absolute error. Our empirical results illustrate that under a LS or a TS there are no significant differences among the three estimators and the most interesting results are obtained in the presence of AOs. The results are complemented by a real data example.

AB - We study robust estimation of a log-linear Poisson model for count time series analysis. More specifically, we study robust versions of maximum likelihood estimators (MLEs) under three different forms of interventions: additive outliers (AOs), transient shifts (TSs) and level shifts (LSs). We estimate the parameters using the MLE, the conditionally unbiased bounded-influence estimator and the Mallows quasi-likelihood estimator and compare all three estimators in terms of their mean-square error, bias and mean absolute error. Our empirical results illustrate that under a LS or a TS there are no significant differences among the three estimators and the most interesting results are obtained in the presence of AOs. The results are complemented by a real data example.

KW - autocorrelation

KW - canonical link

KW - conditionally unbiased bounded-influence estimator

KW - interventions

KW - log-linear Poisson model

KW - Mallows quasi-likelihood estimator

KW - tuning constant

KW - simulation

U2 - 10.1080/00949655.2015.1035271

DO - 10.1080/00949655.2015.1035271

M3 - Journal article

VL - 86

SP - 740

EP - 755

JO - Journal of Statistical Computation and Simulation

JF - Journal of Statistical Computation and Simulation

SN - 0094-9655

IS - 4

ER -