^{a}School of Economics, Sungkyunkwan University, Seoul, Korea
^{b}Division of Medical Entomology, Korea National Institute of Health, Osong, Korea
© 2014 Published by Elsevier B.V. on behalf of Korea Centers for Disease Control and Prevention.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/).
Weekly data | ADF | PP | ||||
---|---|---|---|---|---|---|
None | Intercept | Trend | None | Intercept | Trend | |
All mosquitoes | −5.423* | −6.190* | −6.246* | −5.655* | −6.672* | −6.751* |
Culex tritaeniorhynchus | −8.786* | −9.529* | −9.549* | −6.862* | −7.708* | −7.720* |
Ratio | −6.177* | −7.340* | −7.336* | −5.771* | −6.708* | −6.705* |
Variables | Mean | Maximum value | Minimum value | Standard deviation |
---|---|---|---|---|
All mosquitoes | 7846.3(562.9)* | 87,194.0 | 0.0 | 14,635.1 |
Culex tritaeniorhynchus | 1991.9(217.0)* | 58,769.0 | 0.0 | 5641.4 |
Ratio | 10.2(0.7)* | 73.4 | 0.0 | 17.7 |
Variables | All mosquitoes | Culex tritaeniorhynchus | Ratio of Culex tritaeniorhynchus to all mosquitoes | |||
---|---|---|---|---|---|---|
Test method | AIC | SIC | AIC | SIC | AIC | SIC |
Time lag | 10 | 1 | 4 | 4 | 9 | 4 |
Variable | Last year | 2 years ago | 3 years ago | 4 years ago | R^{2} |
---|---|---|---|---|---|
All mosquitoes | 0.435(0.044)** | 0.269(0.054)** | −0.024(0.056) | 0.310(0.055)** | 0.637 |
Japanese encephalitis | 0.138(0.043)** | 0.353(0.043)** | 0.189(0.050)** | 0.068(0.049) | 0.394 |
Ratio of Japanese encephalitis | 0.162(0.044)** | 0.381(0.041)** | 0.381(0.044)** | −0.085(0.045)*** | 0.717 |
In-sample prediction | |||
---|---|---|---|
All mosquitoes | Culex tritaeniorhynchus | Ratio | |
MSE | 8.881 × 10^{7} | 1.851 × 10^{7} | 84.823 |
Out-of-sample prediction | |||
---|---|---|---|
All mosquitoes | Culex tritaeniorhynchus | Ratio of Culex tritaeniorhynchus | |
MSPE (RO) | 2.045 × 10^{8} | 8.852 × 10^{7} | 86.069 |
MSPE (AD) | 5.229 × 10^{7} | 3.476 × 10^{7} | 78.351 |
*Significant at the 1% level.
ADF = augmented Dickey–Fuller test; PP = Phillips–Perron test.
The number in parenthesis represent standard deviation.
* Significant at the 1% level.
AIC = Akaike information criterion; SIC = Schwarz information criteria.
The results of the autoregressive model showed no differences in superiority with that of autoregressive–moving-average model.
The results of the weekly data were similar to that of the monthly data. Thus, the weekly data, which consists of more data, is described.
* Significant at
** Significant at
*** Significant at
MSE = mean square error.
AD = adding regression; MSPE = mean-square prediction error; RO = rolling regression.