Adaptive order selection for autoregressive models

Chun Shu Chen, Yun Huan Lee, Hung Wei Hsu

研究成果: Article

1 引文 (Scopus)

摘要

Autoregressive model is a popular method for analysing the time dependent data, where selection of order parameter is imperative. Two commonly used selection criteria are the Akaike information criterion (AIC) and the Bayesian information criterion (BIC), which are known to suffer the potential problems regarding overfit and underfit, respectively. To our knowledge, there does not exist a criterion in the literature that can satisfactorily perform under various situations. Therefore, in this paper, we focus on forecasting the future values of an observed time series and propose an adaptive idea to combine the advantages of AIC and BIC but to mitigate their weaknesses based on the concept of generalized degrees of freedom. Instead of applying a fixed criterion to select the order parameter, we propose an approximately unbiased estimator of mean squared prediction errors based on a data perturbation technique for fairly comparing between AIC and BIC. Then use the selected criterion to determine the final order parameter. Some numerical experiments are performed to show the superiority of the proposed method and a real data set of the retail price index of China from 1952 to 2008 is also applied for illustration.

原文English
頁(從 - 到)1963-1974
頁數12
期刊Journal of Statistical Computation and Simulation
84
發行號9
DOIs
出版狀態Published - 2014 九月

指紋

Order Selection
Bayesian Information Criterion
Akaike Information Criterion
Autoregressive Model
Order Parameter
Perturbation techniques
Data Perturbation
Time series
Potential Problems
Unbiased estimator
Perturbation Technique
Dependent Data
Prediction Error
Mean Squared Error
Forecasting
China
Degree of freedom
Numerical Experiment
Experiments
Akaike information criterion

All Science Journal Classification (ASJC) codes

  • Statistics and Probability
  • Modelling and Simulation
  • Statistics, Probability and Uncertainty
  • Applied Mathematics

引用此文

Chen, Chun Shu ; Lee, Yun Huan ; Hsu, Hung Wei. / Adaptive order selection for autoregressive models. 於: Journal of Statistical Computation and Simulation. 2014 ; 卷 84, 編號 9. 頁 1963-1974.
@article{dca0ddd86c54485a8e8462cb176cccb2,
title = "Adaptive order selection for autoregressive models",
abstract = "Autoregressive model is a popular method for analysing the time dependent data, where selection of order parameter is imperative. Two commonly used selection criteria are the Akaike information criterion (AIC) and the Bayesian information criterion (BIC), which are known to suffer the potential problems regarding overfit and underfit, respectively. To our knowledge, there does not exist a criterion in the literature that can satisfactorily perform under various situations. Therefore, in this paper, we focus on forecasting the future values of an observed time series and propose an adaptive idea to combine the advantages of AIC and BIC but to mitigate their weaknesses based on the concept of generalized degrees of freedom. Instead of applying a fixed criterion to select the order parameter, we propose an approximately unbiased estimator of mean squared prediction errors based on a data perturbation technique for fairly comparing between AIC and BIC. Then use the selected criterion to determine the final order parameter. Some numerical experiments are performed to show the superiority of the proposed method and a real data set of the retail price index of China from 1952 to 2008 is also applied for illustration.",
author = "Chen, {Chun Shu} and Lee, {Yun Huan} and Hsu, {Hung Wei}",
year = "2014",
month = "9",
doi = "10.1080/00949655.2013.776559",
language = "English",
volume = "84",
pages = "1963--1974",
journal = "Journal of Statistical Computation and Simulation",
issn = "0094-9655",
publisher = "Taylor and Francis Ltd.",
number = "9",

}

Adaptive order selection for autoregressive models. / Chen, Chun Shu; Lee, Yun Huan; Hsu, Hung Wei.

於: Journal of Statistical Computation and Simulation, 卷 84, 編號 9, 09.2014, p. 1963-1974.

研究成果: Article

TY - JOUR

T1 - Adaptive order selection for autoregressive models

AU - Chen, Chun Shu

AU - Lee, Yun Huan

AU - Hsu, Hung Wei

PY - 2014/9

Y1 - 2014/9

N2 - Autoregressive model is a popular method for analysing the time dependent data, where selection of order parameter is imperative. Two commonly used selection criteria are the Akaike information criterion (AIC) and the Bayesian information criterion (BIC), which are known to suffer the potential problems regarding overfit and underfit, respectively. To our knowledge, there does not exist a criterion in the literature that can satisfactorily perform under various situations. Therefore, in this paper, we focus on forecasting the future values of an observed time series and propose an adaptive idea to combine the advantages of AIC and BIC but to mitigate their weaknesses based on the concept of generalized degrees of freedom. Instead of applying a fixed criterion to select the order parameter, we propose an approximately unbiased estimator of mean squared prediction errors based on a data perturbation technique for fairly comparing between AIC and BIC. Then use the selected criterion to determine the final order parameter. Some numerical experiments are performed to show the superiority of the proposed method and a real data set of the retail price index of China from 1952 to 2008 is also applied for illustration.

AB - Autoregressive model is a popular method for analysing the time dependent data, where selection of order parameter is imperative. Two commonly used selection criteria are the Akaike information criterion (AIC) and the Bayesian information criterion (BIC), which are known to suffer the potential problems regarding overfit and underfit, respectively. To our knowledge, there does not exist a criterion in the literature that can satisfactorily perform under various situations. Therefore, in this paper, we focus on forecasting the future values of an observed time series and propose an adaptive idea to combine the advantages of AIC and BIC but to mitigate their weaknesses based on the concept of generalized degrees of freedom. Instead of applying a fixed criterion to select the order parameter, we propose an approximately unbiased estimator of mean squared prediction errors based on a data perturbation technique for fairly comparing between AIC and BIC. Then use the selected criterion to determine the final order parameter. Some numerical experiments are performed to show the superiority of the proposed method and a real data set of the retail price index of China from 1952 to 2008 is also applied for illustration.

UR - http://www.scopus.com/inward/record.url?scp=84900791534&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84900791534&partnerID=8YFLogxK

U2 - 10.1080/00949655.2013.776559

DO - 10.1080/00949655.2013.776559

M3 - Article

AN - SCOPUS:84900791534

VL - 84

SP - 1963

EP - 1974

JO - Journal of Statistical Computation and Simulation

JF - Journal of Statistical Computation and Simulation

SN - 0094-9655

IS - 9

ER -