发帖
雷达卡
不好意思，又来这里求助啦，关于时间序列的问题，我用SAS Assisst做时间序列的回归，那个阶数要怎么选呢？实在是搞不太清楚，请各位高手指教！！！谢谢大家啦。
模型的样本量比较小，n=11年的数据，我把p=4，p=3，p=2的回归结果都粘出来了，希望大家多多指导，到底选阶数等于几才对呢？？
P=4
普通最小二乘法 Standard Approx
Variable DF Estimate Error t Value Pr > |t|
Intercept 1 -4.2277 3.4914 -1.21 0.2715
A1 1 0.0393 0.1381 0.28 0.7854
A2 1 -0.1518 0.4095 -0.37 0.7236
A3 1 0.3235 0.2632 1.23 0.2651
A4 1 0.000653 0.001754 0.37 0.7224
The AUTOREG Procedure
The AUTOREG Procedure
Estimates of Autoregressive Parameters
Lag Coefficient Standard Error t Value
1 -0.174470 0.698205 -0.25
2 0.527844 0.708922 0.74
3 -0.016733 0.708922 -0.02
4 0.158172 0.698205 0.23
最大似然法
Maximum Likelihood Estimates
SSE 2.30044E-6 DFE 2
MSE 1.15022E-6 Root MSE 0.00107
SBC . AIC .
Regress R-Square 1.0000 Total R-Square 1.0000
Durbin-Watson 1.3956
Standard Approx
Standard Approx
Variable DF Estimate Error t Value Pr > |t|
Intercept 1 11.3674 1.4056 8.09 0.0149
A1 1 -0.4807 0.0487 -9.87 0.0101
A2 1 -2.2118 0.1658 -13.34 0.0056
A3 1 -0.5629 0.0849 -6.63 0.0220
A4 1 0.001328 0.000189 7.02 0.0197
AR1 1 0.0954 0.0226 4.22 0.0517
AR2 1 1.5117 0.006062 249.37 <.0001
AR3 1 0.0954 0.0226 4.22 0.0517
AR4 1 1.0000 1.6768E-6 596376 <.0001
Autoregressive parameters assumed given.
Autoregressive parameters assumed given.
Standard Approx
Variable DF Estimate Error t Value Pr > |t|
Intercept 1 11.3674 0.1012 112.32 <.0001
A1 1 -0.4807 0.003309 -145.25 <.0001
A2 1 -2.2118 0.0120 -184.74 <.0001
A3 1 -0.5629 0.006411 -87.79 0.0001
A4 1 0.001328 0.0000156 85.30 0.0001
P=3
P=3
The AUTOREG Procedure
Estimates of Autoregressive Parameters
Standard
Lag Coefficient Error t Value
1 -0.176232 0.577314 -0.31
2 0.455756 0.523838 0.87
3 0.011142 0.577314 0.02
ERROR: The estimation did not converge in the maximumterations. Continuing assuming convergence.
ERROR: The estimation did not converge in the maximumterations. Continuing assuming convergence.
Maximum Likelihood Estimates
SSE 0.03358367 DFE 3
MSE 0.01119 Root MSE 0.10580
SBC -10.433101 AIC -13.616263
Regress R-Square 0.9449 Total R-Square 0.9253
Durbin-Watson 0.9849
Standard Approx
Standard Approx
Variable DF Estimate Error t Value Pr > |t|
Intercept 1 -2.7419 3.4501 -0.79 0.4848
A1 1 -0.0361 0.1331 -0.27 0.8037
A2 1 -0.5227 0.4418 -1.18 0.3220
A3 1 0.3297 0.2293 1.44 0.2460
A4 1 -0.000280 0.001119 -0.25 0.8187
AR1 1 -0.2795 0.9078 -0.31 0.7783
AR2 1 0.8904 0.2465 3.61 0.0364
AR3 1 -0.1137 0.7493 -0.15 0.8891
The AUTOREG Procedure
The AUTOREG Procedure
Autoregressive parameters assumed given.
Standard Approx
Variable DF Estimate Error t Value Pr > |t|
Intercept 1 -2.7419 3.2907 -0.83 0.4659
A1 1 -0.0361 0.1265 -0.29 0.7939
A2 1 -0.5227 0.4262 -1.23 0.3076
A3 1 0.3297 0.2167 1.52 0.2256
A4 1 -0.000280 0.001038 -0.27 0.8050
P=2
P=2
The AUTOREG Procedure
Estimates of Autoregressive Parameters
Standard
Lag Coefficient Error t Value
1 -0.181332 0.444534 -0.41
2 0.457777 0.444534 1.03
Algorithm converged.
Maximum Likelihood Estimates
SSE 0.03387738 DFE 4
MSE 0.00847 Root MSE 0.09203
SBC -12.726174 AIC -15.511441
Regress R-Square 0.9494 Total R-Square 0.9246
Durbin-Watson 0.8939
Standard Approx
Standard ApproxVariable DF Estimate Error t Value Pr > |t|
Intercept 1 -2.7559 3.0539 -0.90 0.4179
A1 1 -0.0351 0.1128 -0.31 0.7711
A2 1 -0.5261 0.3914 -1.34 0.2501
A3 1 0.3313 0.2038 1.63 0.1794
A4 1 -0.000294 0.000989 -0.30 0.7809
AR1 1 -0.1914 0.3200 -0.60 0.5819
AR2 1 0.8731 0.1776 4.92 0.0080
The AUTOREG Procedure
The AUTOREG Procedure
Autoregressive parameters assumed given.
Standard Approx
Variable DF Estimate Error t Value Pr > |t|
Intercept 1 -2.7559 2.9229 -0.94 0.3991
A1 1 -0.0351 0.1109 -0.32 0.7674
A2 1 -0.5261 0.3768 -1.40 0.2352
A3 1 0.3313 0.1924 1.72 0.1601
A4 1 -0.000294 0.000967 -0.30 0.7762
谢谢大家啦！！！
[此贴子已经被作者于2008-3-19 22:27:32编辑过]
提升卡
置顶卡
沉默卡
变色卡
抢沙发
千斤顶
显身卡
京公网安备 11010802022788号
