This** preview**
has intentionally

**sections.**

*blurred***to view the full version.**

*Sign up*
**Unformatted text preview: **T— Stat Coeff Std Error Variable
Signif
*Hanhk-l:a:an:*2:an:a:a:=1:at:a:a:ata:asa:It:a:an”:at:Han:a:a:an:a:atata:at:at:a****$********M***H**** ********** 95173290697 00248469366 383.03333 1. Constant
000000000
2. TIME
0.00000000 00021697501 00001583668 13.70079 -3.18716
-3.14582 AIC
SIC Fit and assess the standard linear, quadratic, and log linear trend models. Choose model based on AIC and SIC.
AIC is small the better. In linear trend model, AIC is 16.73805. In quadratic trend model,AIC is 16.73103. In log linear trend model,AIC is -3.18716.
SIC is small the better. In linear trend model, SIC is 16.77940. In quadratic trend
model, SIC is 16.79304. In log linear trend model, SIC is -3.l4582. Therefore, linear trend is perfect model in this case. b.
WW a 60000
7 50000
'-- 40000 30000 . .m Milliulilrllui ‘0
. l A ALWJI lam “gr 0"" ¥~,.x‘¢L4IJULMV Y - 20000
. Mil-H .0 r H W "V i
wii' W U! V“
2 V 10000
3 40000 0
'D
'6 30000
£
20000
10000 i
0 , PmlHllenlﬁJV The curves above are atmost as the same shape as the curves below.
Despite the outliers. the curve is smooth. Therefore. there is no obvious effect. and it does not inﬂuence the trend estimate VCI‘) much. f—J 15000 - -- 12500
10000 -
7500
5000
2500 -
1' 1—37 “14'— 5 “e y 0 Estimated Seasonal Factors: hits The columns are almost at the same height. Therefore. there is no effect. soooo
5000C K
4ouoo I
30000 | zoeoo “I h" I I H
,IIIII H.mllrhlhll I'Ltiﬁhrtftll , 10000 Jan Feb Ma: Apr May Jun Jul Aug Sep Oct Nov Dec Hits with Quadratic Trend Forecast ...

View
Full Document

- Fall '11
- JeanFrancisLamarche
- Economics, Approximation, Derivative, Linear Trend Model, log linear trend