|
INVESTIGATION OF ORDER SELECTION IN THE VECTOR AUTOREGRESSION MODEL
|
|
|
D.Petrusevich
|
|
|
||
|
|
|
|
1314-2704
|
|
|
||
|
English
|
|
|
20
|
|
|
2.1
|
|
|
|
|
|
||
|
The econometrical time series are under investigation in this paper. The mathematical
models describing connections between two time series in the form of VAR(p), Vector Autoregression, models are investigated. Their connection is confirmed by means of cointegration tests. If connected time series are described with some models (here the ARIMA, Autoregressive integrated moving average, models) with certain lag orders it?s assumed that maximal orders of ARIMA models are the upper limits for the orders of the VAR model. This point is investigated in the computational experiment. Non-linear terms are included in the VAR model. Such models could describe connections between some time series better. The ordinary least squares method is used to evaluate coefficients in the VAR equations. This method allows non-linear terms to be added into equations. Also there are two ways of ARIMA orders selection. The first one uses values of the information criteria (Akaike and Bayes critera), the second one is based on the correlation functions plots analysis. The maxima locations indicate maximal value of orders of the ARIMA model. Thus there are two ways to select VAR model orders. Both of them are tested. There wage index and money income index time series of the Dynamic series of macroeconomic statistics of the Russian Federation are investigated are investigated in the computational experiment. Investigating non-linear models the VAR model with high coefficient of determination is obtained. |
|
|
conference
|
|
|
||
|
||
|
20th International Multidisciplinary Scientific GeoConference SGEM 2020
|
|
|
20th International Multidisciplinary Scientific GeoConference SGEM 2020, 18 - 24 August, 2020
|
|
|
Proceedings Paper
|
|
|
STEF92 Technology
|
|
|
International Multidisciplinary Scientific GeoConference-SGEM
|
|
|
SWS Scholarly Society; Acad Sci Czech Republ; Latvian Acad Sci; Polish Acad Sci; Russian Acad Sci; Serbian Acad Sci & Arts; Natl Acad Sci Ukraine; Natl Acad Sci Armenia; Sci Council Japan; European Acad Sci, Arts & Letters; Acad Fine Arts Zagreb Croatia; C
|
|
|
191-198
|
|
|
18 - 24 August, 2020
|
|
|
website
|
|
|
cdrom
|
|
|
6986
|
|
|
ARIMA; VAR; time series; Akaike information criterion (AIC)
|
|