季節(jié)ARIMA與RegARIMA模型

1、file: 7b2c3file: 1211692file: 7gener1-1file: dummy010.00E+005.00E+101.00E+111.50E+112.00E+112.50E+113.00E+113.50E+114.00E+11808284868890929496980002GDPGDPF10,00020,00030,00040,00050,00060,00070,00080,000929394959697989900010203040506070809GDP-1.00-0.75-0.50-0.250.000.250.500.751.00246810121416182022

2、24AR000100-6-4-20246255075100125150175200Z1（file:arfunc）-1.00-0.75-0.50-0.250.000.250.500.751.0024681012141618202224AR000100（file:arfunc）-6-4-20246255075100125150175200Z1-1.5-1.0-0.50.00.51.01.5-1.5-1.0-0.50.00.51.01.5AR rootsInverse Roots of AR/MA Polynomial(s)xt = 0.8 xt 4+ut 序列序列的相關(guān)圖與偏相關(guān)圖（的相關(guān)圖與偏相

3、關(guān)圖（file:7gener1-text4）z1-6-4-20246255075100125150175200Z1Bxt = - 0.8 xt 4+ut序列的相關(guān)圖與偏相關(guān)圖（序列的相關(guān)圖與偏相關(guān)圖（file:7gener1-text4）-1.5-1.0-0.50.00.51.01.5-1.5-1.0-0.50.00.51.01.5AR rootsInverse Roots of AR/MA Polynomial(s) .) , 2 , 1( ,4 , 0.) , 2 , 1( ,4,)8 . 0(- jjkjjkjkz2（4）自回歸系數(shù)為負(fù)的）自回歸系數(shù)為負(fù)的SAR過程過程z9-1.5-1.

4、0-0.50.00.51.01.5-1.5-1.0-0.50.00.51.01.5AR rootsInverse Roots of AR/MA Polynomial(s)-6-4-20246102030405060708090100Z2Bz2b-5-4-3-2-101234255075100125150175200Z2-1.5-1.0-0.50.00.51.01.5-1.5-1.0-0.50.00.51.01.5MA rootsInverse Roots of AR/MA Polynomial(s)z5-4-3-2-1012345255075100125150175200Z2B-1.5-1.0

5、-0.50.00.51.01.5-1.5-1.0-0.50.00.51.01.5MA rootsInverse Roots of AR/MA Polynomial(s)z5b-10-8-6-4-20246255075100125150175200Z4，phi=0.4, alpha=0.6-0.4-0.20.00.20.40.60.81.024681012141618202224AR100100-0.6-0.4-0.20.00.20.40.60.81.024681012141618202224ar100100-6-4-202468102030405060708090100Z6-1.5-1.0-0

6、.50.00.51.01.5-1.5-1.0-0.50.00.51.01.5AR rootsInverse Roots of AR/MA Polynomial(s)z6-1.5-1.0-0.50.00.51.01.5-1.5-1.0-0.50.00.51.01.5AR rootsInverse Roots of AR/MA Polynomial(s)SAR(1,0,0)(1,0,0)12序列，序列， (1-0.5L)(1- 0.9L12) z4t = ut的的相關(guān)圖相關(guān)圖呈拖尾特征。而偏自相關(guān)函數(shù)在呈拖尾特征。而偏自相關(guān)函數(shù)在k=13以以后呈截尾特征。后呈截尾特征。z4-6-4-2024625

7、5075100125150175200Z4505)1)(1 (4)1 (3)1)(1 (201)1 (212111211212111211kkkkkkk-4-3-2-101234102030405060708090100Z7-1.5-1.0-0.50.00.51.01.5-1.5-1.0-0.50.00.51.01.5MA rootsInverse Roots of AR/MA Polynomial(s)z7SMA(0,0,1)(0,0,1)12序列序列 xt = (1+ 0.4 L)(1+ 0.6 L12)ut 的相關(guān)圖與偏相關(guān)圖的相關(guān)圖與偏相關(guān)圖-1.5-1.0-0.50.00.51.01

8、.5-1.5-1.0-0.50.00.51.01.5MA rootsInverse Roots of AR/MA Polynomial(s)-3-2-101234255075100125150175200Z3z3總結(jié)總結(jié):-.3-.2-.1.0.1.2787980818283848586878889DSDLNY-1.5-1.0-0.50.00.51.01.5-1.5-1.0-0.50.00.51.01.5MA rootsInverse Roots of AR/MA Polynomial(s)100200300400500600700800900787980818283848586878889Y

9、YF-.3-.2-.1.0.1.2787980818283848586878889DSDLNY-1.5-1.0-0.50.00.51.01.5-1.5-1.0-0.50.00.51.01.5AR rootsMA rootsInverse Roots of AR/MA Polynomial(s)預(yù)測(cè)評(píng)價(jià)預(yù)測(cè)評(píng)價(jià)-1.5-1.0-0.50.00.51.01.5-1.5-1.0-0.50.00.51.01.5MA rootsInverse Roots of AR/MA Polynomial(s)-.4-.2.0.2.4.6246810121419601965197019751980198519901

10、9952000ResidualActualFittedEViews操作：點(diǎn)擊操作：點(diǎn)擊View，選，選residual Diagnostics，Serial Correlation LM Testvt的相關(guān)圖、偏相關(guān)圖如下，的相關(guān)圖、偏相關(guān)圖如下，vt中已不存在自相關(guān)。中已不存在自相關(guān)。010000200003000040000500006000070000800001960 1965 1970 1975 1980 1985 1990 1995 2000YYFYF1010,00020,00030,00040,00050,00060,00070,00080,000010,00030,00050,

11、00070,00090,000GDPYF1YFY0.20.40.60.81.01.21.41.61.8IIIIIIIVIIIIIIIVIIIIIIIVIII1999200020012002Y-.03-.02-.01.00.01.02.030.00.40.81.21.62.0IIIIIIIVIIIIIIIVIIIIIIIVIII1999200020012002ResidualActualFitted-1.5-1.0-0.50.00.51.01.5-1.5-1.0-0.50.00.51.01.5AR rootsInverse Roots of AR/MA Polynomial(s)0.20.40.

12、60.81.01.21.41.61.8IIIIIIIVIIIIIIIVIIIIIIIVIII1999200020012002YYF01102030405060708090100P2殘差圖合格殘差圖合格01199920002001200220032004200520062007D2 去除不顯著虛擬變量，再次回歸，得，去除不顯著虛擬變量，再次回歸，得，觀察殘差的相關(guān)圖和偏相關(guān)圖。觀察殘差的相關(guān)圖和偏相關(guān)圖。-80-4004080-400-300-200-1000100200199920002001200220032004200520062007ResidualActualFitted-4-20246810255075100125150175200Y1-4-2