No abstract

<p>Researchers have been looking for methods to prevent, control and provide lifelong protection to humans against dengue disease which is brought by the dengue-carrying mosquito called the Aedes Aegypti. However, such prevention, control and protection will best be aided by a dengue case prediction model. This study used the Negative Binomial Regression to forecast the dengue case incidence in Metro Manila, Philippines using principal components as explanatory variables. To ensure that the dengue cases are predictable, close returns plot (CRP) was performed.   The logarithm of dengue case incidence which were assigned as response variables have showed higher value of variance over the mean which validates the use of negative binomial regression. Principal Component Analysis utilizing Nino 3.4 sea surface temperature (SST), precipitation and minimum temperature was used in the study. The acquired principal components (PC1, PC2, PC3 and PC4) were used as the explanatory variables for the negative binomial regression to calculate the number of the logarithm of dengue case incidence. However, to improve the calculated value of DHF cases in comparison to its actual value, residuals from the negative binomial regression were treated using moving average approach. The data used in this study were from 1994-2010 climatological data. Results for both negative binomial and moving average were combined to get the forecasted dengue incidence. Forecasted values showed a maximum of 12% difference from the actual DHF cases indicating a high forecasting performance. This study which focused on predicting the possible dengue incidence in the central districts of the Philippines  is believed to be essential to create plans of action to prevent and control this disease.</p>

No abstract

No abstract

No abstract