Лабораторная работа №4
1. Статическая регрессия(У с Х)
Dependent Variable: Y | ||||
Method: Least Squares | ||||
Date: 10/19/08 Time: 11:40 | ||||
Sample: 1996:1 2007:1 | ||||
Included observations: 41 | ||||
Variable | Coefficient | Std. Error | t-Statistic | Prob. |
C | -109.6260 | 26.40488 | -4.151731 | 0.0002 |
X | 2.394248 | 0.794963 | 3.011773 | 0.0045 |
R-squared | 0.188696 | Mean dependent var | -179.9293 | |
Adjusted R-squared | 0.167894 | S. D. dependent var | 86.63545 | |
S. E. of regression | 79.02874 | Akaike info criterion | 11.62505 | |
Sum squared resid | 243576.1 | Schwarz criterion | 11.70864 | |
Log likelihood | -236.3135 | F-statistic | 9.070779 | |
Durbin-Watson stat | 0.776619 | Prob(F-statistic) | 0.004542 |
2. Авторегрессия(У с У(-1))
Dependent Variable: Y | ||||
Method: Least Squares | ||||
Date: 10/19/08 Time: 11:41 | ||||
Sample(adjusted): 1996:2 2007:1 | ||||
Included observations: 40 after adjusting endpoints | ||||
Variable | Coefficient | Std. Error | t-Statistic | Prob. |
C | -40.75285 | 18.77519 | -2.170570 | 0.0363 |
Y(-1) | 0.797677 | 0.095068 | 8.390631 | 0.0000 |
R-squared | 0.649455 | Mean dependent var | -182.6875 | |
Adjusted R-squared | 0.640230 | S. D. dependent var | 85.89665 | |
S. E. of regression | 51.52150 | Akaike info criterion | 10.77058 | |
Sum squared resid | 100869.7 | Schwarz criterion | 10.85503 | |
Log likelihood | -213.4116 | F-statistic | 70.40269 | |
Durbin-Watson stat | 2.024539 | Prob(F-statistic) | 0.000000 |
3. Модель опережающего показателя(У с Х(-1))
Dependent Variable: Y | ||||
Method: Least Squares | ||||
Date: 10/19/08 Time: 11:41 | ||||
Sample(adjusted): 1996:2 2007:1 | ||||
Included observations: 40 after adjusting endpoints | ||||
Variable | Coefficient | Std. Error | t-Statistic | Prob. |
C | -122.0796 | 28.25104 | -4.321242 | 0.0001 |
X(-1) | 2.134081 | 0.886500 | 2.407310 | 0.0210 |
R-squared | 0.132324 | Mean dependent var | -182.6875 | |
Adjusted R-squared | 0.109490 | S. D. dependent var | 85.89665 | |
S. E. of regression | 81.05794 | Akaike info criterion | 11.67691 | |
Sum squared resid | 249674.8 | Schwarz criterion | 11.76136 | |
Log likelihood | -231.5382 | F-statistic | 5.795142 | |
Durbin-Watson stat | 0.674811 | Prob(F-statistic) | 0.021034 |
4. Модель скорости роста – нельзя построить для имеющихся TS рядов, можно только для DS)
5. Модель распределенных запаздываний(У с Х Х(-1))
Dependent Variable: Y | ||||
Method: Least Squares | ||||
Date: 10/19/08 Time: 11:42 | ||||
Sample(adjusted): 1996:2 2007:1 | ||||
Included observations: 40 after adjusting endpoints | ||||
Variable | Coefficient | Std. Error | t-Statistic | Prob. |
C | -101.1511 | 29.69302 | -3.406562 | 0.0016 |
X | 1.782648 | 0.972176 | 1.833667 | 0.0748 |
X(-1) | 1.016638 | 1.054163 | 0.964403 | 0.3411 |
R-squared | 0.204605 | Mean dependent var | -182.6875 | |
Adjusted R-squared | 0.161610 | S. D. dependent var | 85.89665 | |
S. E. of regression | 78.65009 | Akaike info criterion | 11.63993 | |
Sum squared resid | 228876.0 | Schwarz criterion | 11.76660 | |
Log likelihood | -229.7987 | F-statistic | 4.758871 | |
Durbin-Watson stat | 0.725815 | Prob(F-statistic) | 0.014481 |
6. Модель частичной корректировки(У с У(-1) Х)
Dependent Variable: Y | ||||
Method: Least Squares | ||||
Date: 10/19/08 Time: 11:42 | ||||
Sample(adjusted): 1996:2 2007:1 | ||||
Included observations: 40 after adjusting endpoints | ||||
Variable | Coefficient | Std. Error | t-Statistic | Prob. |
C | -45.25015 | 20.01451 | -2.260868 | 0.0297 |
Y(-1) | 0.846748 | 0.119561 | 7.082116 | 0.0000 |
X | -0.447789 | 0.653534 | -0.685181 | 0.4975 |
R-squared | 0.653847 | Mean dependent var | -182.6875 | |
Adjusted R-squared | 0.635136 | S. D. dependent var | 85.89665 | |
S. E. of regression | 51.88496 | Akaike info criterion | 10.80797 | |
Sum squared resid | 99605.83 | Schwarz criterion | 10.93464 | |
Log likelihood | -213.1595 | F-statistic | 34.94463 | |
Durbin-Watson stat | 2.044538 | Prob(F-statistic) | 0.000000 |
7. Фальстарт или приведенная форма(У с У(-1) Х(-1))
Dependent Variable: Y | ||||
Method: Least Squares | ||||
Date: 10/19/08 Time: 11:43 | ||||
Sample(adjusted): 1996:2 2007:1 | ||||
Included observations: 40 after adjusting endpoints | ||||
Variable | Coefficient | Std. Error | t-Statistic | Prob. |
C | -37.33795 | 21.47071 | -1.739018 | 0.0903 |
Y(-1) | 0.782748 | 0.105669 | 7.407536 | 0.0000 |
X(-1) | 0.213779 | 0.626307 | 0.341332 | 0.7348 |
R-squared | 0.650556 | Mean dependent var | -182.6875 | |
Adjusted R-squared | 0.631667 | S. D. dependent var | 85.89665 | |
S. E. of regression | 52.13108 | Akaike info criterion | 10.81744 | |
Sum squared resid | 100553.1 | Schwarz criterion | 10.94410 | |
Log likelihood | -213.3488 | F-statistic | 34.44117 | |
Durbin-Watson stat | 2.001214 | Prob(F-statistic) | 0.000000 |
8. Авторегрессионные ошибки(У с У(-1) Х Х(-1))
Dependent Variable: Y | ||||
Method: Least Squares | ||||
Date: 10/19/08 Time: 11:43 | ||||
Sample(adjusted): 1996:2 2007:1 | ||||
Included observations: 40 after adjusting endpoints | ||||
Variable | Coefficient | Std. Error | t-Statistic | Prob. |
Y(-1) | 0.928577 | 0.113684 | 8.168076 | 0.0000 |
X | -0.546664 | 0.756288 | -0.722826 | 0.4743 |
X(-1) | 0.980807 | 0.675624 | 1.451706 | 0.1550 |
R-squared | 0.627257 | Mean dependent var | -182.6875 | |
Adjusted R-squared | 0.607109 | S. D. dependent var | 85.89665 | |
S. E. of regression | 53.84089 | Akaike info criterion | 10.88198 | |
Sum squared resid | 107257.1 | Schwarz criterion | 11.00865 | |
Log likelihood | -214.6396 | Durbin-Watson stat | 2.113156 |
Выводы:
Все константы с кроме модели «Фальстарт или приведенная форма» получились значимые. R-squared колеблется от 0.132 до 0.654, самая близкой к реальности моделью получилась «Модель частичной корректировки», хотя она все равно слабо отражает действительность, т. к. R-squared должет приближаться к единице.