2. 35 1 - 2557 16 : 2 7 ... 28 39 15 3. 65.77 , , . , . , . ( )
4. 2 16 - 2557 2 j ( 1) 1 :
http://www.na.fs.fed.us/spfo/pubs/uf/techguide/basic.htm 5. 3 16 -
2557 1 air pruning ( 2) 1. air pruning 2. 4.5 x 14 950- 1,000 7.3
25 7.3 25 2 6. 4 16 - 2557 1 (air pruning) air pruning air pruning
7. 5 16 - 2557 820 750 20 3. 4. ( 3) 4.1 2 8 air pruning 4.2
(air-pruning) 2 4.3 3 4-8 air pruning 5. 3 8. 6 16 - 2557 7 air
pruning ( 4) Jaenicke, H. 1999. Good Tree Nursery Practices :
Practical Guidelines for Research Nurseries. International Centre
for Research in Agroforestry. Nairobi, Kenya. 93 p. Landis, T.D.
1990. Containers: types and functions. Pages 1-39. In: T.D. Landis,
R.W. Tinus, S.E. McDonald and J.P. Barnett (eds.). The Container
Tree Nursery Manual, Volume 2. Agric. Handbk. 674. Washington
DC:U.S. Department of Agriculture, Forest Service. Mydin, K.K.,
T.A. Soman and J. Jacob. 2010. Root trainer technique in rubber:
Its modern, cost-effective and labour-friendly. Rubber Asia
(Jul.-Aug.): 57-61. Nelson, W.R. 1996. Butt sweep reduced with good
container design. Australian Forest Growers Biennial Conference,
Mount Gambier. 2 p. Soman, T.A. and C.K. Saraswathy Amma. 1999.
Root trainer nursery for Hevea. Indian Journal of Natural Rubber
Research 12: 17-22. 4 2 . . 9. 7 16 - 2557 1. 2. 12% 12% (total
porosity) 50-80% (Jaenicke, 1999) 3. 4. 5. - 6. (C.E.C.) 7. 8. 10.
8 16 - 2557 pH (Mydin et al., 2010; John and Matthan, 2012) 3 1
(pH) (Total porosity) (Aeration porosity) (Water-holding porosity)
(Dry bulk density) 1. (pH) pH pH 0.5 1.0 pH 4.5-5.5 pH 6.0 2.
(total porosity) 50 Havis and Hamilton (1976) 1 pH 4.2-6 6-7 6-7
11. 9 16 - 2557 (total porosity) 50% (aeration porosity) 20-25%
Whitcomb (1988) aeration po- rosity 25-35% (aeration porosity)
(water-holding porosity) aeration porosity 13% 3. (bulk density) (
) 1.0-1.8 20-30 . 6 ( 2) 4.5 x 14 4 3 (Dick- son et al., 1960;
Haase, 2008) ( 4) 1. 2 6 pH1 Total porosity2 (%) Aeration porosity3
(%) Water-holding porosity4 (%) Dry bulk density5 (g/cm3 ) : (3:1)
5.22 5.88 45.29 20.76 24.53 0.59 :: 5.89 5.31 43.86 17.52 26.34
0.60 (3:1:1) :: 5.16 5.86 61.48 35.86 25.62 0.74 (1:1:1) :: 5.38
5.81 59.48 20.57 38.91 0.54 (2:1:1) :: 5.80 5.86 58.00 20.10 37.90
0.51 : (2:1:1:1) :: 5.30 5.44 55.10 30.19 24.91 0.69 (2:1:1) 4 1 ,
2 , 3 , 4 , 5 12. 10 16 - 2557 ::: (2:1:1:1) ( 3 .) : (3:1) ::
(3:1:1) 2. (shoot : root ratio) 2:1 6:1 3 : (shoot : root) 2:1-3:1
2:1 6:1 (Dickson Quality Index) (Dickson et al., 1960) () (.) (.) +
() () DQI = 13. 11 16 - 2557 shoot : root 3 ::: (2:1:1:1) : (3:1)
:: (3:1:1) : (3:1) shoot : root shoot : root 2 (air pruning) shoot
: root ( 1) 4 4 (.) (.) () () : (.) quality index : (3:1) 6.03b
77.93a 6.48c 2.66b 2.45cd 0 0.59bc :: 5.97b 73.58b 8.25ab 2.19cd
3.93a 0 0.64b (3:1:1) :: 5.49c 61.51d 4.86d 2.34c 2.13e 0 0.54c
(1:1:1) :: 5.11d 62.37d 4.21de 1.94d 2.21de 0 0.43d (2:1:1) ::
6.51a 78.14a 8.91a 2.84b 3.20b 0 0.77a : (2:1:1:1) :: 5.25d 58.50e
3.89e 2.04cd 1.94e 0 0.45d (2:1:1) 5.80b 65.54c 8.02b 3.17a 2.57c
32.54 0.81a C.V. (%) 2.76 2.96 7.16 8.59 7.05 7.32 DMRT 95% 14. 12
16 - 2557 3. 32.54 . 4. ::: (2:1:1:1) : (3:1) : : (3:1:1) ( 2) 5. :
(3:1) :: (3:1:1) shoot : root 6. Dickson quality index ( 5) 1 (1-6)
(7) 4 1. : (3:1) 2. :: (3:1:1) 3. :: (1:1:1) 4. :: (2:1:1) 5. :::
(2:1:1:1) 6. :: (2:1:1) 7. 15. 13 16 - 2557 2 () () 2:1:1:1 3:1
3:1:1 Dickson, A., A.L. Leaf and J.F. Hosner. 1960. Quality
appraisal of white spruce and white pine seedling stock in
nurseries. Forestry Chronicle 36: 10-13. Haase, D.L. 2008.
Understanding forest seedling quality: measurements and
interpretation. Tree Planters Notes. 52(2): 24-30. Havis J.R. and
W.W. Hamilton. 1976. Physical 16. 14 16 - 2557 properties of
container media. Journal of Arboliculture 2(7): 139-140. Jaenicke,
H. 1999. Good Tree Nursery Practices: Practical Guidelines for
Research Nurseries. International Centre for Reseach in
Agroforestry, Nairobi, Kenya. 93 p. John, J. and R.K. Matthan.
2012. The root trainer technique, a novel method of propagation of
rubber. The Rubber International Magazine 14(11): 68-71. Mydin,
K.K., T.A. Soman and J. Jacob. 2010. Root trainer technique in
rubber: Its modern, cost-effective and labour-friendly. Rubber Asia
(Jul.-Aug.): 57-61. Whitcomb, C.E. 1988. Plant production in
containers. Stillwater, OK: Lacebark Publications. 633 p. 5 4 N (%)
P (%) K (%) Ca (%) Mg (%) S (%) Fe (./ .) Mn (./ .) Zn (./ .) :
(3:1) 3.08 0.33 1.52 1.57 0.16 0.21 236 89 25 :: 3.72 0.46 1.57
1.02 0.29 0.33 182 93 34 (3:1:1) :: 3.15 0.45 1.90 1.36 0.17 0.28
162 97 30 (1:1:1) :: 3.23 0.35 1.50 1.57 0.17 0.24 202 66 27
(2:1:1) :: 3.25 0.45 1.48 0.98 0.30 0.24 156 48 34 : (2:1:1:1) ::
3.19 0.45 1.88 1.62 0.16 0.26 188 96 26 (2:1:1) 3.06 0.30 1.44 1.25
0.17 0.24 345 75 30 17. 15 16 - 2557 .. 2552 2556 2 1) 3 3 5 2) 3)
1. 3 1.1 3 2 2548 20 2555 1,163 18. 16 16 - 2557 (Stationary)
(Mean) (Variance) (Covariance) Unit root Augmented Dickey- Fuller
test (ADF-test)
-----------------------------------------------------------------------------------------------------------------------
1.1 3 2 2548 20 2555 1,163 (Stationary) (Mean) (Variance)
(Covariance) Unit root Augmented Dickey-Fuller test (ADF-test) Xt =
Xt-1 + t (random walk process) Xt = + Xt-1 + t (random walk with
drift) Xt = + t + Xt-1 + t (random walk with drift and linear time
trend) H0 : = 0 Ha : 0 H0 Xt (Nonstationary) 1.2 4 Box and Jenkins,
Hybrid Forecasting, Multiple Regression Combine Forecasting
Multiple Regression Intervention Generalized Autoregressive
Conditional Heteroscedasticity (GARCH) 1.2.1 1.2.1.1 Box and
Jenkins (random walk process)
-----------------------------------------------------------------------------------------------------------------------
3 2 2548 20 2555 1,163 (Stationary) (Mean) (Variance) (Covariance)
Unit root Augmented Dickey-Fuller test (ADF-test) Xt = Xt-1 + t
(random walk process) Xt = + Xt-1 + t (random walk with drift) Xt =
+ t + Xt-1 + t (random walk with drift and linear time trend) H0 :
= 0 Ha : 0 H0 Xt (Nonstationary) 1.2 4 Box and Jenkins, Hybrid
Forecasting, Multiple Regression Combine Forecasting Multiple
Regression Intervention Generalized Autoregressive Conditional
Heteroscedasticity (GARCH) 1.2.1 1.2.1.1 Box and Jenkins (ran-dom
walk with drift)
-----------------------------------------------------------------------------------------------------------------------
1.1 3 2 2548 20 2555 1,163 (Stationary) (Mean) (Variance)
(Covariance) Unit root Augmented Dickey-Fuller test (ADF-test) Xt =
Xt-1 + t (random walk process) Xt = + Xt-1 + t (random walk with
drift) Xt = + t + Xt-1 + t (random walk with drift and linear time
trend) H0 : = 0 Ha : 0 H0 Xt (Nonstationary) 1.2 4 Box and Jenkins,
Hybrid Forecasting, Multiple Regression Combine Forecasting
Multiple Regression Intervention Generalized Autoregressive
Conditional Heteroscedasticity (GARCH) 1.2.1 1.2.1.1 Box and
Jenkins (random walk with drift and linear time trend)
-----------------------------------------------------------------------------------------------------------------------
3 2 2548 20 2555 1,163 (Stationary) (Mean) (Variance) (Covariance)
Unit root Augmented Dickey-Fuller test (ADF-test) Xt = Xt-1 + t
(random walk process) Xt = + Xt-1 + t (random walk with drift) Xt =
+ t + Xt-1 + t (random walk with drift and linear time trend) H0 :
= 0 Ha : 0 H0 Xt (Nonstationary) 1.2 4 Box and Jenkins, Hybrid
Forecasting, Multiple Regression Combine Forecasting Multiple
Regression Intervention Generalized Autoregressive Conditional
Heteroscedasticity (GARCH) 1.2.1 1.2.1.1 Box and Jenkins ,
-----------------------------------------------------------------------------------------------------------------------
2 20 2555 1,163 (Stationary) (Mean) (Variance) (Covariance) Unit
root Augmented Dickey-Fuller test (ADF-test) Xt = Xt-1 + t (random
walk proces Xt = + Xt-1 + t (random walk with drift) Xt = + t +
Xt-1 + t (random walk with d linear time trend) H0 : = 0 Ha : 0 H0
Xt (Nonstationary) 1.2 4 Box Jenkins, Hybrid Forecasting, Multiple
Regression Combine Forecasting M Regression Intervention
Generalized Autoregressive Con Heteroscedasticity (GARCH) 1.2.1
1.2.1.1 Box and Jenkins H0 Xt (Nonstationary) 1.2 4 Box Jenkins,
Hybrid Forecasting, Multiple Regression Combine Forecasting
Multiple Regression Intervention Generalized Autoregressive
Conditional Heteroscedasticity (GARCH) 1.2.1 1.2.1.1 Box Jenkins
ARIMA ARIMA SARIMA (Seasonal Integrated Auto Regressive and Moving
Average) 3 Seasonal AutoRegressive SAR : (P), Integrated (I)
Seasonal Moving Average SMA : (Q) SAR (P) Xt Box and Jenkins ARIMA
ARIMA SARIMA (Seasonal Integrated AutoRegressive and Moving
Average) 3 Seasonal AutoRegressive SAR : (P), Integrated (I)
Seasonal Moving Average SMA : (Q) SAR (P) Xt XtS ,..,XtPS MA (Q) Xt
tS ,, tPS Integrated (I) difference) SARIMA (p,d,q)(P,D,Q)S
(1-1B--pBp )(1-1Bs --PBPs )(1-B)d (1-Bs )D Xt = +(1-1B--qBq )(1-1Bs
--QBQs )t P = autoregressive d = q = moving average P =
autoregressive D = s Q = moving average S = B = backshift operator
1.2.1.2 Hybrid Forecasting Box and Jenkins Artificial Neural
Network (ANN) Box and Jenkins Artificial Neural Network Back
propagation algorithm MA (Q) Xt Box and Jenkins ARIMA ARIMA SARIMA
(Seasonal Integrated AutoRegressive and Moving Average) 3 Seasonal
AutoRegressive SAR : (P), Integrated (I) Seasonal Moving Average
SMA : (Q) SAR (P) Xt XtS ,..,XtPS MA (Q) Xt tS ,, tPS Integrated
(I) difference) SARIMA (p,d,q)(P,D,Q)S (1-1B--pBp )(1-1Bs --PBPs
)(1-B)d (1-Bs )D Xt = +(1-1B--qBq )(1-1Bs --QBQs )t P =
autoregressive d = q = moving average P = autoregressive D = s Q =
moving average S = B = backshift operator 1.2.1.2 Hybrid
Forecasting Box and Jenkins Artificial Neural Network (ANN) Box and
Jenkins Artificial Neural Network Back propagation algorithm
Artificial Neural Network 3 1) input Integrated (I) (difference)
SARIMA (p,d,q)(P,D,Q)S Box and Jenkins ARIMA ARIMA SARIMA (Seasonal
Integrated AutoRegressive and Moving Average) 3 Seasonal
AutoRegressive SAR : (P), Integrated (I) Seasonal Moving Average
SMA : (Q) SAR (P) Xt XtS ,..,XtPS MA (Q) Xt tS ,, tPS Integrated
(I) difference) SARIMA (p,d,q)(P,D,Q)S (1-1B--pBp )(1-1Bs --PBPs
)(1-B)d (1-Bs )D Xt = +(1-1B--qBq )(1-1Bs --QBQs )t P =
autoregressive d = q = moving average P = autoregressive D = s Q =
moving average S = B = backshift operator 1.2.1.2 Hybrid
Forecasting Box and Jenkins Artificial Neural Network (ANN) Box and
Jenkins Artificial Neural Network Back propagation algorithm
Artificial Neural Network 3 1) input 2) Hidden Input Output Input P
= autoregressive d = q = moving average P = autoregressive D = sQ =
moving average S = B = backshift operator 1.2.1.2 Hybrid
Forecasting Box Jenkins Artificial Neural Network (ANN) Box Jenkins
Artificial Neural Network Back propagation algorithm 19. 17 16 -
2557 Artificial Neural Network 3 1) Input 2) Hidden Input Output
Input Output 3) Output Artificial Neural Network (, 2555) 1.2.1.3
Multiple Regression 3 (Dummy Variable) Intervention (Ordinary Least
Squares : OLS) GARCH Intervention Dummy Intervention 2 (Step
Function) (Pulse Function) ( , 2552) Dummy Intervention It = 1
Intervention, It = 0 Intervention Generalized Autoregressive
Conditional Heteroscedasticity (GARCH) stochastic variable (error
term) Engle, Robert F. ( , 2555) ARCH Bollerslev GARCH (Conditional
Variance) t Conditional Variance
-----------------------------------------------------------------------------------------------------------------------
Intervention Dummy Intervention 2 (Step Function) (Pulse Function)
(, 2552) Dummy Intervention It = 1 Intervention It = 0 Intervention
GeneralizedAutoregressiveConditionalHeteroscedasticity(GARCH)
stochastic variable (error term) Engle, Robert F. (, 2555) ARCH
1986 Bollerslev GARCH (Conditional Variance) t Conditional Variance
Et-12 t = 2 t = 0 +i2 t-1+i2 t-i Conditional Variance t 2 t t = t2
t t = 2 v = 1 1.2.1.4 Combine Forecasting Combine Forecasting F1t 1
F2t 2 ef1 ef2 Combine Forecasting CFt = wF1t + (1-w)F2t q i=1 i=1 q
Conditional Variance
-----------------------------------------------------------------------------------------------------------------------
Intervention Dummy Intervention 2 (Step Function) (Pulse Function)
(, 2552) Dummy Intervention It = 1 Intervention It = 0 Intervention
GeneralizedAutoregressiveConditionalHeteroscedasticity(GARCH)
stochastic variable (error term) Engle, Robert F. (, 2555) ARCH
1986 Bollerslev GARCH (Conditional Variance) t Conditional Variance
Et-12 t = 2 t = 0 +i2 t-1+i2 t-i Conditional Variance t 2 t t = t2
t t = 2 v = 1 1.2.1.4 Combine Forecasting Combine Forecasting F1t 1
F2t 2 ef1 ef2 Combine Forecasting CFt = wF1t + (1-w)F2t q i=1 i=1 q
-----------------------------------------------------------------------------------------------------------------------
Intervention Dummy Intervention 2 (Step Function) (Pulse Function)
(, 2552) Dummy Intervention It = 1 Intervention It = 0 Intervention
GeneralizedAutoregressiveConditionalHeteroscedasticity(GARCH)
stochastic variable (error term) Engle, Robert F. (, 2555) ARCH
1986 Bollerslev GARCH (Conditional Variance) t Conditional Variance
Et-12 t = 2 t = 0 +i2 t-1+i2 t-i Conditional Variance t 2 t t = t2
t t = 2 v = 1 1.2.1.4 Combine Forecasting Combine Forecasting F1t 1
F2t 2 ef1 ef2 Combine Forecasting CFt = wF1t + (1-w)F2t q i=1 i=1 q
-----------------------------------------------------------------------------------------------------------------------
Intervention Dum Intervention 2 Function) 2552) Dummy It = 1
Intervention It = 0 Intervention GeneralizedAutoregressiveC
stochastic v (, 2555) GARCH (Conditional Variance) t Conditional
Variance Et-12 t = 2 t = 0 +i2 t-1 Conditional Variance t t = t2 t
1.2.1.4 Combine Forecasting Combine Forecasting F1t 1 F2 ef1 ef2
Combine Forecasting CFt = wF1t + (1-w)F2t q i=1 ervention Dummy
vention 2 (Step (Pulse Function) (, Dummy Intervention ntervention
Intervention
GeneralizedAutoregressiveConditionalHeteroscedasticity(GARCH)
stochastic variable (error term) Engle, Robert F. , 2555) ARCH 1986
Bollerslev GARCH t Conditional Variance Et-12 t = 2 t = 0 +i2
t-1+i2 t-i tional Variance t 2 t 2 t t = 2 v = 1 mbine Forecasting
asting 1 F2t 2 ef1 ef2 ne Forecasting wF1t + (1-w)F2t q i=1 i=1 q
1.2.1.4 Combine Forecasting F1t 1 F2t 2 ef1 ef2 Combine Forecasting
: CFt = wF1t + (1-w)F2t Combine Forecasting : eCFt = (CFt -Yt+1 ) =
weF1t + (1-w)ef2t 20. 18 16 - 2557 Combine Forecasting
-----------------------------------------------------------------------------------------------------------------------
Combine Forecasting eCFt = (CFt-Yt+1) = weF1t + (1-w)ef2t Combine
Forecasting 2 eCFt = w2 2 eF1t + (1-w)2 2 eF2t +2w(1-w)eF1t eF2t 2
eF1t 1 2 eF2t 2 eF1t eF2t 1 2 (w) W = 2 eF2t - eF1t eF2t eF1t eF2t
1 eF1teF2t W = 2 eF2t - eF1t eF2t : (2548) 1.2.2 ( Mean Absolute
Error: MAE) (Mean Absolute Percentage Error : MAPE) (Root Mean
Square Error : RMSE ) MAE = MAPE = N k+n t n t=k k+n n ()2 t=k 2
eF1t + 2 eF2t - 2eF1t eF2t ^ ^ ^ 2 eF1t + 2 eF2t - 2eF1t eF2t ^^ ^
k+n n t=k Yt |t| X 100
-----------------------------------------------------------------------------------------------------------------------
Combine Forecasting eCFt = (CFt-Yt+1) = weF1t + (1-w)ef2t Combine
Forecasting 2 eCFt = w2 2 eF1t + (1-w)2 2 eF2t +2w(1-w)eF1t eF2t 2
eF1t 1 2 eF2t 2 eF1t eF2t 1 2 (w) W = 2 eF2t - eF1t eF2t eF1t eF2t
1 eF1teF2t W = 2 eF2t - eF1t eF2t : (2548) 1.2.2 ( Mean Absolute
Error: MAE) (Mean Absolute Percentage Error : MAPE) (Root Mean
Square Error : RMSE ) MAE = MAPE = N k+n t n t=k k+n n ()2 t=k 2
eF1t + 2 eF2t - 2eF1t eF2t ^ ^ ^ 2 eF1t + 2 eF2t - 2eF1t eF2t ^^ ^
k+n n t=k Yt |t| X 100 1 Combine Forecasting eCFt = (CFt-Yt+1) =
weF1t + (1-w)ef2t Combine Forecasting 2 eCFt = w2 2 eF1t + (1-w)2 2
eF2t +2w(1-w)eF1t eF2t t 1 t 2 1t eF2t 1 2 (w) W = 2 eF2t - eF1t
eF2t eF1t eF2t 1 eF1teF2t W = 2 eF2t - eF1t eF2t (2548) : MAE)
(Mean Error : MAPE) or : RMSE ) MAE = MAPE = N k+n t n t=k k+n n
()2 t=k 2 eF1t + 2 eF2t - 2eF1t eF2t ^ ^ ^ 2 eF1t + 2 eF2t - 2eF1t
eF2t ^^ ^ k+n n t=k Yt |t| X 100 2 ombine Forecasting = weF1t +
(1-w)ef2t Combine Forecasting t + (1-w)2 2 eF2t +2w(1-w)eF1t eF2t 1
2 2 w) eF2t - eF1t eF2t 1 eF1teF2t 2 eF2t - eF1t eF2t (Mean MAE =
MAPE = N k+n t n t=k k+n n ()2 t=k 2 eF2t - 2eF1t eF2t ^ ^ + 2 eF2t
- 2eF1t eF2t ^^ k+n n t=k Yt |t| X 100 1 2 (w)
-----------------------------------------------------------------------------------------------------------------------
Combine Forecasting eCFt = (CFt-Yt+1) = weF1t + (1-w)ef2t Combine
Forecasting 2 eCFt = w2 2 eF1t + (1-w)2 2 eF2t +2w(1-w)eF1t eF2t 2
eF1t 1 2 eF2t 2 eF1t eF2t 1 2 (w) W = 2 eF2t - eF1t eF2t eF1t eF2t
1 eF1teF2t W = 2 eF2t - eF1t eF2t : (2548) 1.2.2 ( Mean Absolute
Error: MAE) (Mean Absolute Percentage Error : MAPE) (Root Mean
Square Error : RMSE ) MAE = MAPE = N k+n t n t=k k+n n ()2 t=k 2
eF1t + 2 eF2t - 2eF1t eF2t ^ ^ ^ 2 eF1t + 2 eF2t - 2eF1t eF2t ^^ ^
k+n n t=k Yt |t| X 100
----------------------------------------------- Combine Forecasting
eCFt = (CFt-Yt+1) = weF1t + (1-w)ef2t Combine Forecasting 2 eCFt =
w2 2 eF1t + (1-w)2 2 eF2t +2w(1-w)eF1t eF2t 2 eF1t 1 2 eF2t 2 eF1t
eF2t 1 2 (w) W = 2 eF2t - eF1t eF2t eF1t eF2t 1 eF1teF2t W = 2 eF2t
- eF1t eF2t (2548) 1.2.2 Absolute Error: MAE) (Mean te Percentage
Error : MAPE) Mean Square Error : RMSE ) MAE = MAPE = N k+n t n t=k
k+n n ()2 t=k 2 eF1t + 2 eF2t - 2eF1t eF2t ^ ^ ^ 2 eF1t + 2 eF2t -
2eF1t eF2t ^^ ^ k+n n t=k Yt |t| X 100
-----------------------------------------------------------------------------------------------------------------------
Combine Forecasting eCFt = (CFt-Yt+1) = weF1t + (1-w)ef2t Combine
Forecasting 2 eCFt = w2 2 eF1t + (1-w)2 2 eF2t +2w(1-w)eF1t eF2t 2
eF1t 1 2 eF2t 2 eF1t eF2t 1 2 (w) W = 2 eF2t - eF1t eF2t eF1t eF2t
1 eF1teF2t W = 2 eF2t - eF1t eF2t : (2548) 1.2.2 ( Mean Absolute
Error: MAE) (Mean Absolute Percentage Error : MAPE) (Root Mean
Square Error : RMSE ) MAE = MAPE = N k+n t n t=k k+n n ()2 t=k 2
eF1t + 2 eF2t - 2eF1t eF2t ^ ^ ^ 2 eF1t + 2 eF2t - 2eF1t eF2t ^^ ^
k+n n t=k Yt |t| X 100
-----------------------------------------------------------------------------------------------------------------------
Combine Forecasting eCFt = (CFt-Yt+1) = weF1t + (1-w)ef2t Combine
Forecasting 2 eCFt = w2 2 eF1t + (1-w)2 2 eF2t +2w(1-w)eF1t eF2t 2
eF1t 1 2 eF2t 2 eF1t eF2t 1 2 (w) W = 2 eF2t - eF1t eF2t eF1t eF2t
1 eF1teF2t W = 2 eF2t - eF1t eF2t : (2548) 1.2.2 ( Mean Absolute
Error: MAE) (Mean Absolute Percentage Error : MAPE) (Root Mean
Square Error : RMSE ) MAE = MAPE = N k+n t n t=k k+n n ()2 t=k 2
eF1t + 2 eF2t - 2eF1t eF2t ^ ^ ^ 2 eF1t + 2 eF2t - 2eF1t eF2t ^^ ^
k+n n t=k Yt |t| X 100 : (2548) 1.2.2 (Mean Absolute Error : MAE)
(Mean Absolute Percentage Error : MAPE) (Root Mean Square Error :
RMSE )
-----------------------------------------------------------------------------------------------------------------------
Combine Forecasting eCFt = (CFt-Yt+1) = weF1t + (1-w)ef2t Combine
Forecasting 2 eCFt = w2 2 eF1t + (1-w)2 2 eF2t +2w(1-w)eF1t eF2t 2
eF1t 1 2 eF2t 2 eF1t eF2t 1 2 (w) W = 2 eF2t - eF1t eF2t eF1t eF2t
1 eF1teF2t W = 2 eF2t - eF1t eF2t : (2548) 1.2.2 ( Mean Absolute
Error: MAE) (Mean Absolute Percentage Error : MAPE) (Root Mean
Square Error : RMSE ) MAE = MAPE = N k+n t n t=k k+n n ()2 t=k 2
eF1t + 2 eF2t - 2eF1t eF2t ^ ^ ^ 2 eF1t + 2 eF2t - 2eF1t eF2t ^^ ^
k+n n t=k Yt |t| X 100
-----------------------------------------------------------------------------------------------------------------------
Combine Forecasting eCFt = (CFt-Yt+1) = weF1t + (1-w)ef2t Combine
Forecasting 2 eCFt = w2 2 eF1t + (1-w)2 2 eF2t +2w(1-w)eF1t eF2t 2
eF1t 1 2 eF2t 2 eF1t eF2t 1 2 (w) W = 2 eF2t - eF1t eF2t eF1t eF2t
1 eF1teF2t W = 2 eF2t - eF1t eF2t : (2548) 1.2.2 ( Mean Absolute
Error: MAE) (Mean Absolute Percentage Error : MAPE) (Root Mean
Square Error : RMSE ) MAE = MAPE = N k+n t n t=k k+n n ()2 t=k 2
eF1t + 2 eF2t - 2eF1t eF2t ^ ^ ^ 2 eF1t + 2 eF2t - 2eF1t eF2t ^^ ^
k+n n t=k Yt |t| X 100
-----------------------------------------------------------------------------------------------------------------------
Combine Forecasting eCFt = (CFt-Yt+1) = weF1t + (1-w)ef2t Combine
Forecasting 2 eCFt = w2 2 eF1t + (1-w)2 2 eF2t +2w(1-w)eF1t eF2t 2
eF1t 1 2 eF2t 2 eF1t eF2t 1 2 (w) W = 2 eF2t - eF1t eF2t eF1t eF2t
1 eF1teF2t W = 2 eF2t - eF1t eF2t : (2548) 1.2.2 ( Mean Absolute
Error: MAE) (Mean Absolute Percentage Error : MAPE) (Root Mean
Square Error : RMSE ) MAE = MAPE = N k+n t n t=k k+n n ()2 t=k 2
eF1t + 2 eF2t - 2eF1t eF2t ^ ^ ^ 2 eF1t + 2 eF2t - 2eF1t eF2t ^^ ^
k+n n t=k Yt |t| X 100 RMSE = 0 21. 19 16 - 2557 2. 2.1 3
(Purchasing Managers Index:PMI) 2546 2556 120 2.2 (order of
integration) 2.2.1 (Stationary) Unit root Augmented Dickey-Fuller
test (ADF-test) 2.2.2 (Cointegration) (Nonstation- ary) (Long-run
equilibrium relationship) Cointegration Cointegration Cointegration
Johansen Juselius 2.2.3 (Ordinary least square : OLS) Stock Watson
Dynamic ordinary least square : DOLS Co-integration ( , 2554) Stock
and Watson 1993 Dynamic ordinary least square : DOLS Co-integration
(, 2554) Qt tM' i i i Pti i i i Yti iAti i i Et M = [c, , , ] , X =
[1, P, YA] m , n l The lengths of leads and lags of the regressors
: Ahmed Al-Azzam and David Hawdon (1999) 2.2.4 (Simultaneous
Equation System) D = a0 + a1P + a2Y + u1 S = b0 + b1P + b2W + u2 D
= S Stock and Watson 1993 Dynamic ordinary least square : DOLS
Co-integration (, 2554) Qt tM' i i i Pti i i i Yti iAti i i Et M =
[c, , , ] , X = [1, P, YA] m , n l The lengths of leads and lags of
the regressors : Ahmed Al-Azzam and David Hawdon (1999) 2.2.4
(Simultaneous Equation System) D = a0 + a1P + a2Y + u1 S = b0 + b1P
+ b2W + u2 D = S m , n l The lengths of leads and lags of the
regressors : Ahmed Al-Azzam and David Hawdon (1999) 2.2.4
(Simultaneous Equation System) D = a0 + a1 P + a2 Y + u1 S = b0 +
b1 P + b2 W + u2 D = S 22. 20 16 - 2557
-----------------------------------------------------------------------------------------------------------------------
D = a0 + a1P + a2Y + u1 S = b0 + b1P + b2W + u2 D = S P = a0b0 b1
a1 a2 b1 a1 Y b2 b1 a1 1 = u1u2 b1 a1 : (2544) 3. : (2544) 3.
(www.rubberthaiforward.com) (www.rubberthai.com) 1. 3 3 Box Jenkins
ARIMA 3 Multiple Regression 2 2548 20 2555 1,163 (Mean Absolute
Error : MAE) (Mean Absolute Percentage Error : MAPE) (Root Mean
Square Error : RMSE ) 5 1.1 Box Jenkins Unit root 3 ADF-test 3 I(1)
3 (RSS3) SARIMA (1,1,0)(1,0,1)5 backshift operator
-----------------------------------------------------------------------------------------------------------------------
(www.rubberthaiforward.com) (www.rubberthai.com) 1. 3 3 Box and
Jenkins ARIMA 3 Multiple Regression 2 2548 20 2555 1,163 ( Mean
Absolute Error : MAE) (Mean Absolute Percentage Error : MAPE) (Root
Mean Square Error : RMSE ) 5 1.1 Box and Jenkins Unit root 3
ADF-test 3 I(1) 3 (RSS3) SARIMA (1,1,0)(1,0,1)5 backshift operator
(1-0.417B)(1-0.644B5 )(1-B)lnYt = (1-0.636B5 )t yt = 0.417yt-1 -
0.644yt-5 + t + 0.636t-5 Stationary R2 = 0.667 R2 = 0.998 B
backshift operator (BJ Xt = Xt-J) Yt 3 (RSS3) 1.2 Hybrid
Forecasting B backshift operator
-----------------------------------------------------------------------------------------------------------------------
(www.rubberthaiforward.com) (www.rubberthai.com) 1. 3 3 Box and
Jenkins ARIMA 3 Multiple Regression 2 2548 20 2555 1,163 ( Mean
Absolute Error : MAE) (Mean Absolute Percentage Error : MAPE) (Root
Mean Square Error : RMSE ) 5 1.1 Box and Jenkins Unit root 3
ADF-test 3 I(1) 3 (RSS3) SARIMA (1,1,0)(1,0,1)5 backshift operator
(1-0.417B)(1-0.644B5 )(1-B)lnYt = (1-0.636B5 )t yt = 0.417yt-1 -
0.644yt-5 + t + 0.636t-5 Stationary R2 = 0.667 R2 = 0.998 B
backshift operator (BJ Xt = Xt-J) Yt 3 (RSS3) 1.2 Hybrid
Forecasting Yt 3 (RSS3) 23. 21 16 - 2557 1.2 Hybrid Forecasting
ARIMA ANN ANN Input 2 (node) SARIMA(1,1,0)(1,0,1)5 Hidden 1 2
Output 1 sum of squares error 1.3 Multiple Regression 3 (RSS3)
(TOCOM) (SICOM) (SHFE) (EX) Intervention (I) Shock 2 Hamberger
Crisis GARCH (Ordinary Least Squares : OLS)
-----------------------------------------------------------------------------------------------------------------------
ANN Input 2 SARIMA(1,1,0)(1,0,1)5 Hidden 1 2 Output 1 sum of
squares error 1.3 Multiple Regression 3 (RSS3) (TOCOM) (SICOM)
(SHFE) (EX) Intervention (I) Shock 2 Hamberger Crisis GARCH
(Ordinary Least Squares : OLS) LOG(RSS3) = 0.0001 + 0.085LOG(TOCOM)
+ 0.049LOG(TOCOMt-1) z-statistic (9.315) (3.985)
+0.246LOG(SICOM)+0.207LOG(SICOMt-1)+0.252LOG(SHFE) (13.747)
(10.370) (13.804) 0.459LOG (EX) - 0.271I + 0.249I1 (-4.667)
(-39.361) (19.979) GARCH = 0.00001 + 0.2592 + 0.685GARCHt-1
z-statistic (8.055) (21.777) Stationary R2 = 0.696 D-W = 2.106 F
statistic 210.266 1.4 Combine Forecasting Hybrid Forecasting
Multiple Regression F1 Hybrid Forecasting F2 Multiple Regression
(W) t-1 1.4 Combine Forecasting Hybrid Forecasting Multiple
Regression F1 Hybrid Forecasting F2 Multiple Regression (W) WF1 =
0.195 WF2 = 0.804 Combine Forecasting (CF) 3 CFt = 0.195F1t +
0.804F2t 1.5 MAE MAPE RMSE 4 Hybrid Forecasting Combine
Forecasting, Multiple Regression SARIMA 1 23-27 2555 Hybrid
Forecasting 2 24. 22 16 - 2557 1 4 Model MAE MAPE RMSE SARIMA(
1,1,0)(1,0,1)5 0.970 1.061 1.353 Hybrid Forecasting 0.214 0.668
0.988 Multiple Regression 0.786 0.842 1.339 Combine Forecasting
0.068 0.072 1.008 2 3 // RSS3 SARIMA (1,1,0)(1,0,1)5 Hybrid
Forecasting Multiple Regression Combine Forecasting 23/7/2555 91.59
98.64 90.99 92.90 92.53 24/7/2555 91.59 90.12 90.97 90.77 90.81
25/7/2555 90.89 91.59 90.25 90.30 90.29 26/7/2555 91.40 90.66 90.78
90.37 90.45 27/7/2555 92.25 91.61 91.67 91.38 91.46 : / . 2. (DNR)
(Pw) (PMI_ch) (PMI_ja) (SNR) (Stock) 2.1 (Unit root test) Augmented
Dickey-Fuller test 3 ADF-test ( 3) 2.2 (cointegration test)
Johensen - Juselius Trace Statistic 4 Rank Trace Statistic Trace
Statistic 25. 23 16 - 2557 4 Eigen value Trace statistic 5% 1% *
95% , ** 99%
-----------------------------------------------------------------------------------------------------------------------
3 ADP - stat 1% 5% 10% DNR -3.460* -4.041 -3.450 -3.150 (1) Pw
-4.540** -4.047 -3.453 -3.152 (2) PMI_ch -3.596* -4.050 -3.454
-3.152 (0) PMI_ja -8.418** -4.037 -3.448 -3.149 (1) SNR -3.883**
-4.037 -3.448 -3.149 (1) STOCK -4.009** -4.042 -3.450 -3.150 (1) *
: 95% , ** : 99% ADF-test ( 3) 2.2 (cointegration test) Johensen -
Juselius (1990) Trace Statistic 4 Eigen value Trace statistic 5% 1%
r= 0 ** > 0 0.208 64.595 55.245 62.521 r 1 ** r > 1 0.156
37.708 35.010 41.081
-----------------------------------------------------------------------------------------------------------------------
3 ADP - stat 1% 5% 10% DNR -3.460* -4.041 -3.450 -3.150 (1) Pw
-4.540** -4.047 -3.453 -3.152 (2) PMI_ch -3.596* -4.050 -3.454
-3.152 (0) PMI_ja -8.418** -4.037 -3.448 -3.149 (1) SNR -3.883**
-4.037 -3.448 -3.149 (1) STOCK -4.009** -4.042 -3.450 -3.150 (1) *
: 95% , ** : 99% ADF-test ( 3) 2.2 (cointegration test) Johensen -
Juselius (1990) Trace Statistic 4 Eigen value Trace statistic 5% 1%
r= 0 ** > 0 0.208 64.595 55.245 62.521 r 1 ** r > 1 0.156
37.708 35.010 41.081
-----------------------------------------------------------------------------------------------------------------------
3 ADP - stat 1% 5% 10% DNR -3.460* -4.041 -3.450 -3.150 (1) Pw
-4.540** -4.047 -3.453 -3.152 (2) PMI_ch -3.596* -4.050 -3.454
-3.152 (0) PMI_ja -8.418** -4.037 -3.448 -3.149 (1) SNR -3.883**
-4.037 -3.448 -3.149 (1) STOCK -4.009** -4.042 -3.450 -3.150 (1) *
: 95% , ** : 99% ADF-test ( 3) 2.2 (cointegration test) Johensen -
Juselius (1990) Trace Statistic 4 Eigen value Trace statistic 5% 1%
r= 0 ** > 0 0.208 64.595 55.245 62.521 r 1 ** r > 1 0.156
37.708 35.010 41.081
-----------------------------------------------------------------------------------------------------------------------
3 ADP - stat 1% 5% 10% DNR -3.460* -4.041 -3.450 -3.150 (1) Pw
-4.540** -4.047 -3.453 -3.152 (2) PMI_ch -3.596* -4.050 -3.454
-3.152 (0) PMI_ja -8.418** -4.037 -3.448 -3.149 (1) SNR -3.883**
-4.037 -3.448 -3.149 (1) STOCK -4.009** -4.042 -3.450 -3.150 (1) *
: 95% , ** : 99% ADF-test ( 3) 2.2 (cointegration test) Johensen -
Juselius (1990) Trace Statistic 4 Eigen value Trace statistic 5% 1%
r= 0 ** > 0 0.208 64.595 55.245 62.521 r 1 ** r > 1 0.156
37.708 35.010 41.081
-----------------------------------------------------------------------------------------------------------------------
3 ADP - stat 1% 5% 10% DNR -3.460* -4.041 -3.450 -3.150 (1) Pw
-4.540** -4.047 -3.453 -3.152 (2) PMI_ch -3.596* -4.050 -3.454
-3.152 (0) PMI_ja -8.418** -4.037 -3.448 -3.149 (1) SNR -3.883**
-4.037 -3.448 -3.149 (1) STOCK -4.009** -4.042 -3.450 -3.150 (1) *
: 95% , ** : 99% ADF-test ( 3) 2.2 (cointegration test) Johensen -
Juselius (1990) Trace Statistic 4 Eigen value Trace statistic 5% 1%
r= 0 ** > 0 0.208 64.595 55.245 62.521 r 1 ** r > 1 0.156
37.708 35.010 41.081
-----------------------------------------------------------------------------------------------------------------------
3 ADP - stat 1% 5% 10% DNR -3.460* -4.041 -3.450 -3.150 (1) Pw
-4.540** -4.047 -3.453 -3.152 (2) PMI_ch -3.596* -4.050 -3.454
-3.152 (0) PMI_ja -8.418** -4.037 -3.448 -3.149 (1) SNR -3.883**
-4.037 -3.448 -3.149 (1) STOCK -4.009** -4.042 -3.450 -3.150 (1) *
: 95% , ** : 99% ADF-test ( 3) 2.2 (cointegration test) Johensen -
Juselius (1990) Trace Statistic 4 Eigen value Trace statistic 5% 1%
r= 0 ** > 0 0.208 64.595 55.245 62.521 r 1 ** r > 1 0.156
37.708 35.010 41.081 r 2 * r > 2 0.110 18.928 18.397 23.152 r 3
** r > 3 0.046 5.447 3.841 6.634 * : 95% , ** : 99% 4 Rank Trace
Statistic Trace Statistic r 0 2.3 Johensen - Juselius (1990) Trace
Statistic 5 Eigenvalue Trace statistic 5% 1% r= 0 ** r > 0 0.234
52.745 35.010 32.064 r 2 * r > 2 0.110 18.928 18.397 23.152 r 3
** r > 3 0.046 5.447 3.841 6.634 * : 95% , ** : 99% 4 Rank Trace
Statistic Trace Statistic r 0 2.3 Johensen - Juselius (1990) Trace
Statistic 5 Eigenvalue Trace statistic 5% 1% r= 0 ** r > 0 0.234
52.745 35.010 32.064 r 2 * r > 2 0.110 18.928 18.397 23.152 r 3
** r > 3 0.046 5.447 3.841 6.634 * : 95% , ** : 99% 4 Rank Trace
Statistic Trace Statistic r 0 2.3 Johensen - Juselius (1990) Trace
Statistic 5 Eigenvalue Trace statistic 5% 1% r= 0 ** r > 0 0.234
52.745 35.010 32.064 r 2 * r > 2 0.110 18.928 18.397 23.152 r 3
** r > 3 0.046 5.447 3.841 6.634 * : 95% , ** : 99% 4 Rank Trace
Statistic Trace Statistic r 0 2.3 Johensen - Juselius (1990) Trace
Statistic 5 Eigenvalue Trace statistic 5% 1% r= 0 ** r > 0 0.234
52.745 35.010 32.064 r 2 * r > 2 0.110 18.928 18.397 23.152 r 3
** r > 3 0.046 5.447 3.841 6.634 * : 95% , ** : 99% 4 Rank Trace
Statistic Trace Statistic r 0 2.3 Johensen - Juselius (1990) Trace
Statistic 5 Eigenvalue Trace statistic 5% 1% r= 0 ** r > 0 0.234
52.745 35.010 32.064 r 2 * r > 2 0.110 18.928 18.397 23.152 r 3
** r > 3 0.046 5.447 3.841 6.634 * : 95% , ** : 99% 4 Rank Trace
Statistic Trace Statistic r 0 2.3 Johensen - Juselius (1990) Trace
Statistic 5 Eigenvalue Trace statistic 5% 1% r= 0 ** r > 0 0.234
52.745 35.010 32.064 r 0 2.3 Johensen - Juselius Trace Statistic 5
Rank Trace Statistic Trace Statistic r 0 3 ADP - stat 1% 5% 10% DNR
-3.460* -4.041 -3.450 -3.150 I(1) Pw -4.540** -4.047 -3.453 -3.152
I(2) PMI_ch -3.596* -4.050 -3.454 -3.152 I(0) PMI_ja -8.418**
-4.037 -3.448 -3.149 I(1) SNR -3.883** -4.037 -3.448 -3.149 I (1)
STOCK -4.009** -4.042 -3.450 -3.150 I(1) * 95% , ** 99% 26. 24 16 -
2557 5 Eigen value Trace statistic 5% 1% * 95% , ** 99%
-----------------------------------------------------------------------------------------------------------------------
r 0 2.3 Johensen - Juselius (1990) Trace Statistic 5 Eigenvalue
Trace statistic 5% 1% r= 0 ** r > 0 0.234 52.745 35.010 32.064 r
1 ** r > 1 0.111 22.050 18.397 16.160 r 2** r > 2 0.070 8.429
3.841 2.705 * : 95% , ** : 99% 5 Rank Trace Statistic Trace
Statistic r 0 2.4 Dynamic Ordinary Least Square : DOLS 2.4.1.
lnDNRt = 10.448 - 4.703 lnPwt - 1.021 lnstockt-1 +0.853 lnPMI_jat
(-8.356)** (-23.690)** (11.884)**
-----------------------------------------------------------------------------------------------------------------------
r 0 2.3 Johensen - Juselius (1990) Trace Statistic 5 Eigenvalue
Trace statistic 5% 1% r= 0 ** r > 0 0.234 52.745 35.010 32.064 r
1 ** r > 1 0.111 22.050 18.397 16.160 r 2** r > 2 0.070 8.429
3.841 2.705 * : 95% , ** : 99% 5 Rank Trace Statistic Trace
Statistic r 0 2.4 Dynamic Ordinary Least Square : DOLS 2.4.1.
lnDNRt = 10.448 - 4.703 lnPwt - 1.021 lnstockt-1 +0.853 lnPMI_jat
(-8.356)** (-23.690)** (11.884)**
-----------------------------------------------------------------------------------------------------------------------
r 0 2.3 Johensen - Juselius (1990) Trace Statistic 5 Eigenvalue
Trace statistic 5% 1% r= 0 ** r > 0 0.234 52.745 35.010 32.064 r
1 ** r > 1 0.111 22.050 18.397 16.160 r 2** r > 2 0.070 8.429
3.841 2.705 * : 95% , ** : 99% 5 Rank Trace Statistic Trace
Statistic r 0 2.4 Dynamic Ordinary Least Square : DOLS 2.4.1.
lnDNRt = 10.448 - 4.703 lnPwt - 1.021 lnstockt-1 +0.853 lnPMI_jat
(-8.356)** (-23.690)** (11.884)**
-----------------------------------------------------------------------------------------------------------------------
r 0 2.3 Johensen - Juselius (1990) Trace Statistic 5 Eigenvalue
Trace statistic 5% 1% r= 0 ** r > 0 0.234 52.745 35.010 32.064 r
1 ** r > 1 0.111 22.050 18.397 16.160 r 2** r > 2 0.070 8.429
3.841 2.705 * : 95% , ** : 99% 5 Rank Trace Statistic Trace
Statistic r 0 2.4 Dynamic Ordinary Least Square : DOLS 2.4.1.
lnDNRt = 10.448 - 4.703 lnPwt - 1.021 lnstockt-1 +0.853 lnPMI_jat
(-8.356)** (-23.690)** (11.884)**
-----------------------------------------------------------------------------------------------------------------------
r 0 2.3 Johensen - Juselius (1990) Trace Statistic 5 Eigenvalue
Trace statistic 5% 1% r= 0 ** r > 0 0.234 52.745 35.010 32.064 r
1 ** r > 1 0.111 22.050 18.397 16.160 r 2** r > 2 0.070 8.429
3.841 2.705 * : 95% , ** : 99% 5 Rank Trace Statistic Trace
Statistic r 0 2.4 Dynamic Ordinary Least Square : DOLS 2.4.1.
lnDNRt = 10.448 - 4.703 lnPwt - 1.021 lnstockt-1 +0.853 lnPMI_jat
(-8.356)** (-23.690)** (11.884)**
-----------------------------------------------------------------------------------------------------------------------
r 0 2.3 Johensen - Juselius (1990) Trace Statistic 5 Eigenvalue
Trace statistic 5% 1% r= 0 ** r > 0 0.234 52.745 35.010 32.064 r
1 ** r > 1 0.111 22.050 18.397 16.160 r 2** r > 2 0.070 8.429
3.841 2.705 * : 95% , ** : 99% 5 Rank Trace Statistic Trace
Statistic r 0 2.4 Dynamic Ordinary Least Square : DOLS 2.4.1.
lnDNRt = 10.448 - 4.703 lnPwt - 1.021 lnstockt-1 +0.853 lnPMI_jat
(-8.356)** (-23.690)** (11.884)** 2.4 Dynamic Ordinary Least Square
: DOLS 2.4.1. lnDNRt = 10.448 - 4.703 lnPwt - 1.021 lnstockt-1
+0.853 lnPMI_jat (-8.356)** (-23.690)** (11.884)** R2 = 0.999 adj
R2 = 0.996 D.W = 1.745 : ** 99% DNR Pw stock PMI_ja 2.4.2. lnSNRt =
6.819 + 8.642lnPwt 59.114lnSNRt-3 - 0.003EXSt (38.990)**
(-78.936)** (-5.884)** R2 = 0.999 adj R2 = 0.998 D.W = 1.925 : **
99% SNR EXS 2.4.3 (Simultaneous-Equation Model) lnPwt = 0.272 +
4.430lnSNRt-3 + 0.0002EXSt - 0.077lnstockt-1 + 0.064 lnPMI_jat 3
27. 25 16 - 2557 3 1(4.430) 3 1% 59.114% 4.430% 1(0.0002) 1 0.003%
0.0002% -1(-0.077) 1% 1.021% 0.077% -1 (-0.064) 1% 0.853% 0.064% 3
3. www.rubberthaiforward.com 2 3 28. 26 16 - 2557 Hybrid
Forecasting Univariate - Multiple Regression Multivariate GARCH 2
Combine 2546 2555 3 . .. (International Rubber Consortium Limited:
IRCO) . GARCH (1, 1). :
http://anchan.lib.ku.ac.th/thai-ciard/handle/009/28760 (31 ..
2555). . 2544. 2 . . 208 . . 2548. . 4(2): 18-23. . 2555. . . . 186
. 2552. 29. 27 16 - 2557 SARIMA Intervention. 11(1): 196-214.
.2554.. 29(2) : 1-34. Ahmed Al-Azzam and David Hawdon. 1999.
Estimating the Demand for Energy in Jordan : A Stock- Watson
Dynamic OLS (DOLS) Approach. Department of Economics University of
Surrey (Online). Available from :
http://www.seec.surrey.ac.uk/Research/SEEDS/SEEDS97.pdf (cited Oct.
5 2013). 30. 28 16 - 2557 (Dipping) (Foaming) (Casting) (Extrusion)
(Latex Profile) (Latex Extrusion) 1 2 1 3 1 2 3 (Heat sensitive)
Polyvi- nyl methyl ether (MRPRA, 1983 Gorton, 1994a) heat sensitive
(Gorton, 1994b Blackley, 1997) Extrusion 31. 29 16 - 2557 (Die) 1.
(Polyvinyl methyl ether, PVME) 2 1, 2 3 3 2 1 2 10 78 3 40 8 2 3 10
3 1 3 3 20 35, 40, 45, 50, 55, 60, 75 80 50 75 3 3 2 1 3 4 5 32. 30
16 - 2557 2 3 3 3 20 50 1 3 2 3 10 78 - 3 40 8 3 20 35, 40, 45, 50,
55, 6 50 75 3 3 20 50 3 2 1 3 4 5 10 78 - 2 40 3 3 20 35, 40, 45,
50, 55, 60, 75 50 75 3 3 50 3 2 1 3 4 5 10 78 - 2 40 3 3 20 35, 40,
45, 50, 55, 60, 75 80 50 75 3 3 3 20 50 3 2 1 3 4 5 4 2 10 78 - 2
40 6 10 90 4 2 33. 31 16 - 2557 2. 2.1 1, 2 3 1 2.1.1 3 10% PVME 4
phr ( 1) 20 35, 40, 45, 50, 55, 60, 75 80 50 75 100 20 (filler) 50%
(Calcium carbonate, CaCO3 ) 5 phr 10% PVME 4 phr ( 1 2) 20 50 4,480
4 40 6 3 10% PVME 4 phr 50% (Calcium carbonate, CaCO3 ) 1, 2, 3, 4
5 phr ( 1 2) 20 50 CaCO3 2 phr 3 50% CaCO3 5 phr 3 50% CaCO3 1, 2,
3 4 phr 50% CaCO3 2 phr 5 1 4 2 5 1 - 2 40 6 10 90 - 2 28 10 34. 32
16 - 2557 1 2 50% (CaCO3) 50% CaCO3 ( 100 , phr) ( 100 , phr) 50%
CaCO3 () () () 1. 60% 100 167 40 2. (Deionized water) 14.3 3. 25%
Emulvin W 0.25 1 4. 40% 2.8 7 30 pH pH 7.5 - 8.0 5. 50% 1.25 2.5 6.
50% 1.0 2 7. 50% 1.0 2 8. 50% 0.5 1 9. 10% (PVME) 4 40 1 1 2 14.4 2
4 14.6 3 6 14.7 4 8 14.8 5 10 15 ( 30 - 50 ) 2.1.2 2 10% PVME 35.
33 16 - 2557 4 phr ( 1) 20 30, 35, 40, 45, 50, 55, 60 70 55 7 10%
PVME 4 phr 8 phr 20 50, 55, 60 55 50 60 10% PVME 4 phr 2.1.3 1 10%
PVME 4 phr ( 1) 20 35 8 1 2.2 2 ( 1.26 ) 10% PVME 3 10% PVME 50% 0,
1, 2, 3, 4 5 phr ( 2.48, 2.40, 2.32, 2.31, 2.39 2.37 ) 2 ISO 2321 :
2006 (Tensile strength) (Elongation at break) 300 (300% Modulus) 7
50% CaCO3 () 50% CaCO3, 5 phr ( 173%) 3 50% 5 phr 2.1.2 2 2 10%
PVME 4 phr ( 1) 20 30, 35, 40, 45, 50, 55, 60 70 55 7 10% PVME 4
phr 8 phr 20 50, 55, 60 55 50 10% PVME 4 phr 2 7 50% CaCO3 () 50%
CaCO3, 5 phr ( 173%) 6 3 50% 5 phr 2.1.2 2 2 10% PVME 4 phr ( 1) 20
30, 35, 40, 45, 50, 55, 60 70 55 7 10% PVME 4 phr 8 phr 20 50, 55,
60 55 50 60 10% PVME 4 phr 2 6 3 50% 5 phr () 50% CaCO3 (), () 50%
CaCO3 5 phr ( 173%) 36. 34 16 - 2557 3 3 50% (50% CaCO3 ) 50% CaCO3
( 100 , phr) 6 3 3 50% (50% CaCO3) 50% CaCO3 ( 100 , phr) 0 () - 1
- 2 3 4 5 6 3 3 50% (50% CaCO3) 50% CaCO3 ( 100 , phr) 0 () - 1 - 2
3 4 5 6 3 3 50% (50% CaCO3) 50% CaCO3 ( 100 , phr) 0 () - 1 - 2 3 4
5 6 3 3 50% (50% CaCO3) 50% CaCO3 ( 100 , phr) 0 () - 1 - 2 3 4 5 6
3 3 50% (50% CaCO3) 50% CaCO3 ( 100 , phr) 0 () - 1 - 2 3 4 5 6 3 3
50% (50% CaCO3) 50% CaCO3 ( 100 , phr) 0 () - 1 - 2 3 4 5 6 3 3 50%
(50% CaCO3) 50% CaCO3 ( 100 , phr) 0 () - 1 - 2 3 4 5 6 3 3 50%
(50% CaCO3) 50% CaCO3 ( 100 , phr) 0 () - 1 - 2 3 4 5 6 3 3 50%
(50% CaCO3) 50% CaCO3 ( 100 , phr) 0 () - 1 - 2 3 4 5 6 3 3 50%
(50% CaCO3) 50% CaCO3 ( 100 , phr) 0 () - 1 - 2 3 4 5 6 3 3 50%
(50% CaCO3) 50% CaCO3 ( 100 , phr) 0 () - 1 - 2 3 4 5 6 3 3 50%
(50% CaCO3) 50% CaCO3 ( 100 , phr) 0 () - 1 - 2 3 4 5 6 3 3 50%
(50% CaCO3) 50% CaCO3 ( 100 , phr) 0 () - 1 - 2 3 4 5 6 3 3 50%
(50% CaCO3) 50% CaCO3 ( 100 , phr) 0 () - 1 - 2 3 4 5 6 3 3 50%
(50% CaCO3) 50% CaCO3 ( 100 , phr) 0 () - 1 - 2 3 4 5 6 3 3 50%
(50% CaCO3) 50% CaCO3 ( 100 , phr) 0 () - 1 - 2 3 4 5 0 () 1 2 3 4
5 - - 500 (500% Modulus) 4 5 3 4 50% 37. 35 16 - 2557 300 500 5 50%
1 2 phr 50% 3, 4 5 phr 50% 50% 50% 50% 1 4 phr 50% 2 4 5 7 2 () 10%
PVME 4 phr 50, 55, 60 55 50 60 10% PVME 4 phr 2 7 2 10% PVME 4 phr
10% PVME 4 phr 2 2 10% PVME 4 phr 8 1 8 2.1.3 1 1 10% PVME 4 phr (
1) 20 35 8 8 1 1 2.2 2 ( 1.26 ) 10% PVME 3 10% PVME 50% 0, 1, 2, 3,
4 5 phr ( 2.48, 2.40, 2.32, 2.31, 2.39 2.37 ) 2 ISO 2321 : 2006
(Tensile strength) (Elongation at break) 300 (300% Modulus) 500
(500% Modulus) 4 5 3 4 50% 300 38. 36 16 - 2557 4 2 3 50% 1, 2, 3,
4 5 phr 50% CaCO3 (phr) (%) (N/mm2 ) (%) 300 (N/mm2 ) 500 (N/mm2 )
2 () 0 1 180 11.0 645 1.20 2.00 2 () 0 1 180 7.0 592 1.04 1.89 3 ()
0 1 173 12.6 692 1.21 2.01 1 1 184 9.5 638 1.25 2.15 1 2 176 11.6
633 1.29 2.35 2 1 165 10.8 637 1.38 2.52 2 2 176 9.6 598 1.43 2.86
3 1 176 12.3 635 1.39 2.69 3 2 165 11.1 643 1.32 2.37 4 1 176 10.8
672 1.33 2.47 4 2 165 11.1 601 1.36 2.91 5 1 176 12.3 658 1.34 2.50
5 2 169 11.0 627 1.33 2.71 3 () 0 1 173 14.4 653 1.30 2.42 1 1 184
12.2 653 1.37 2.40 1 2 176 10.9 628 1.42 2.54 2 1 165 13.5 645 1.48
2.69 2 2 176 11.8 563 1.57 2.86 3 1 176 10.3 668 1.49 2.68 3 2 165
10.9 594 1.54 2.89 4 1 176 11.4 609 1.51 2.97 4 2 165 8.9 614 1.32
2.42 5 1 176 9.1 602 1.38 2.71 5 2 169 8.7 611 1.32 2.40 39. 37 16
- 2557 5 2 3 50% CaCO3 (phr) (%) 300 500 2 0 1 180 -36.4 -8.2 -13.3
-5.5 3 0 1 173 14.3 0.0 7.4 20.4 1 1 184 28.4 0.0 9.6 11.6 1 2 176
-6.0 -0.8 10.1 8.1 2 1 165 25.0 1.3 7.2 6.7 2 2 176 22.9 -5.9 9.8
0.0 3 1 176 -16.3 5.2 7.2 -0.4 3 2 165 -1.8 -7.6 16.7 21.9 4 1 176
5.6 -9.4 13.5 20.2 4 2 165 -19.8 2.2 -2.9 -16.8 5 1 176 -26.0 -8.5
3.0 8.4 5 2 169 -20.9 -2.6 -0.8 -11.4 () 2 3 2 10 78 2 3 40 2 3
(heat sensitivity) 10% (Polyvinylmethylether) 4 phr 25 50 50% 5 phr
1 40. 38 16 - 2557 PVME Blackley, D.C. 1997. Heat-sensitizing
coacervants. Polymer Latices Science and Technology. 2nd ed. Vol.
1. Chapman & Hall. London. pp. 303-351 Gorton, T. 1994a. Latex
tubing. Rubber Products Manufacturing Technology. Edited by A.K.
Bhowmick, M.M. Hall and H.A. Benarey. Marcel Dekker, Inc. New York.
pp. 842 Gorton, T. 1994b. Heat-sensitive Dipping. Rubber Products
Manufacturing Technology. Edited by A.K. Bhowmick, M.M. Hall and
H.A. Benarey. Marcel Dekker, Inc. New York. pp. 832 MRPRA. 1983.
Extruded latex tubing Natural Rubber Technical Information Sheet,
L58 41. 39 16 - 2557 44/2557 10 2557 Biotechnology 26 2556 2/2557
110/2557 28 2557 6/2557 31 2557 398/2557 24 2557 38 285/2557 42. 40
16 - 2557 24 2557 2 40 18 27 17 20 2557 2 40 25 27 2557 11 13
2557
